shylee commited on
Commit
3e05360
·
verified ·
1 Parent(s): cb1c692

Add DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain from 91cf0fc7c553

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +1 -0
  2. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/pretrained_model/config.json +107 -0
  3. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/pretrained_model/model.safetensors +3 -0
  4. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/pretrained_model/train_config.json +217 -0
  5. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/training_state/optimizer_param_groups.json +331 -0
  6. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/training_state/optimizer_state.safetensors +3 -0
  7. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/training_state/rng_state.safetensors +3 -0
  8. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/training_state/scheduler_state.json +15 -0
  9. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/training_state/training_step.json +3 -0
  10. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/pretrained_model/config.json +107 -0
  11. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/pretrained_model/model.safetensors +3 -0
  12. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/pretrained_model/train_config.json +217 -0
  13. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/training_state/optimizer_param_groups.json +331 -0
  14. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/training_state/optimizer_state.safetensors +3 -0
  15. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/training_state/rng_state.safetensors +3 -0
  16. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/training_state/scheduler_state.json +15 -0
  17. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/training_state/training_step.json +3 -0
  18. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/pretrained_model/config.json +107 -0
  19. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/pretrained_model/model.safetensors +3 -0
  20. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/pretrained_model/train_config.json +217 -0
  21. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/training_state/optimizer_param_groups.json +331 -0
  22. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/training_state/optimizer_state.safetensors +3 -0
  23. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/training_state/rng_state.safetensors +3 -0
  24. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/training_state/scheduler_state.json +15 -0
  25. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/training_state/training_step.json +3 -0
  26. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/pretrained_model/config.json +107 -0
  27. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/pretrained_model/model.safetensors +3 -0
  28. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/pretrained_model/train_config.json +217 -0
  29. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/training_state/optimizer_param_groups.json +331 -0
  30. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/training_state/optimizer_state.safetensors +3 -0
  31. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/training_state/rng_state.safetensors +3 -0
  32. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/training_state/scheduler_state.json +15 -0
  33. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/training_state/training_step.json +3 -0
  34. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/pretrained_model/config.json +107 -0
  35. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/pretrained_model/model.safetensors +3 -0
  36. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/pretrained_model/train_config.json +217 -0
  37. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/training_state/optimizer_param_groups.json +331 -0
  38. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/training_state/optimizer_state.safetensors +3 -0
  39. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/training_state/rng_state.safetensors +3 -0
  40. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/training_state/scheduler_state.json +15 -0
  41. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/training_state/training_step.json +3 -0
  42. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/pretrained_model/config.json +107 -0
  43. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/pretrained_model/model.safetensors +3 -0
  44. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/pretrained_model/train_config.json +217 -0
  45. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/training_state/optimizer_param_groups.json +331 -0
  46. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/training_state/optimizer_state.safetensors +3 -0
  47. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/training_state/rng_state.safetensors +3 -0
  48. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/training_state/scheduler_state.json +15 -0
  49. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/training_state/training_step.json +3 -0
  50. DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/400000/pretrained_model/config.json +107 -0
.gitattributes CHANGED
@@ -53,3 +53,4 @@ DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_middletrain/wandb/run-202
53
  DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_shorttrain/wandb/run-20250518_143255-5e56tibb/run-5e56tibb.wandb filter=lfs diff=lfs merge=lfs -text
54
  DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_shorttrain/wandb/run-20250519_150138-5e56tibb/run-5e56tibb.wandb filter=lfs diff=lfs merge=lfs -text
55
  DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_middletrain/wandb/run-20250519_150246-fekph8dz/run-fekph8dz.wandb filter=lfs diff=lfs merge=lfs -text
 
 
53
  DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_shorttrain/wandb/run-20250518_143255-5e56tibb/run-5e56tibb.wandb filter=lfs diff=lfs merge=lfs -text
54
  DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_shorttrain/wandb/run-20250519_150138-5e56tibb/run-5e56tibb.wandb filter=lfs diff=lfs merge=lfs -text
55
  DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_middletrain/wandb/run-20250519_150246-fekph8dz/run-fekph8dz.wandb filter=lfs diff=lfs merge=lfs -text
56
+ DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/wandb/run-20250519_150523-bqr8y636/run-bqr8y636.wandb filter=lfs diff=lfs merge=lfs -text
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/pretrained_model/config.json ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "diffusion",
3
+ "n_obs_steps": 2,
4
+ "normalization_mapping": {
5
+ "VISUAL": "MEAN_STD",
6
+ "STATE": "MIN_MAX",
7
+ "ACTION": "MIN_MAX"
8
+ },
9
+ "input_features": {
10
+ "observation.state": {
11
+ "type": "STATE",
12
+ "shape": [
13
+ 6
14
+ ]
15
+ },
16
+ "observation.images.FrontCam": {
17
+ "type": "VISUAL",
18
+ "shape": [
19
+ 3,
20
+ 480,
21
+ 640
22
+ ]
23
+ },
24
+ "observation.images.TopCam": {
25
+ "type": "VISUAL",
26
+ "shape": [
27
+ 3,
28
+ 480,
29
+ 640
30
+ ]
31
+ },
32
+ "observation.images.WristCam": {
33
+ "type": "VISUAL",
34
+ "shape": [
35
+ 3,
36
+ 480,
37
+ 640
38
+ ]
39
+ }
40
+ },
41
+ "output_features": {
42
+ "action": {
43
+ "type": "ACTION",
44
+ "shape": [
45
+ 6
46
+ ]
47
+ }
48
+ },
49
+ "device": "cuda",
50
+ "use_amp": false,
51
+ "horizon": 16,
52
+ "n_action_steps": 8,
53
+ "drop_n_last_frames": 7,
54
+ "vision_backbone": "resnet50",
55
+ "crop_shape": [
56
+ 224,
57
+ 224
58
+ ],
59
+ "crop_is_random": false,
60
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
61
+ "use_group_norm": false,
62
+ "spatial_softmax_num_keypoints": 32,
63
+ "use_separate_rgb_encoder_per_camera": false,
64
+ "down_dims": [
65
+ 256,
66
+ 512,
67
+ 1024
68
+ ],
69
+ "kernel_size": 5,
70
+ "n_groups": 8,
71
+ "diffusion_step_embed_dim": 128,
72
+ "use_film_scale_modulation": true,
73
+ "noise_scheduler_type": "DDIM",
74
+ "num_train_timesteps": 100,
75
+ "beta_schedule": "squaredcos_cap_v2",
76
+ "beta_start": 0.0001,
77
+ "beta_end": 0.02,
78
+ "prediction_type": "epsilon",
79
+ "clip_sample": true,
80
+ "clip_sample_range": 1.0,
81
+ "num_inference_steps": 10,
82
+ "do_mask_loss_for_padding": false,
83
+ "optimizer_lr": 0.0001,
84
+ "optimizer_betas": [
85
+ 0.95,
86
+ 0.999
87
+ ],
88
+ "optimizer_eps": 1e-08,
89
+ "optimizer_weight_decay": 1e-06,
90
+ "scheduler_name": "cosine",
91
+ "scheduler_warmup_steps": 500,
92
+ "pre_resize_shape": [
93
+ 224,
94
+ 224
95
+ ],
96
+ "freeze_vision_backbone": false,
97
+ "use_ac": false,
98
+ "ac_threshold": 0.99,
99
+ "use_pcg": false,
100
+ "pcg_weight": 0.1,
101
+ "use_bid": false,
102
+ "bid_batch_size": 15,
103
+ "bid_mode_size": 3,
104
+ "bid_temporal_coherence": 0.5,
105
+ "weak_model_path": null,
106
+ "weak_model_config": null
107
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/pretrained_model/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d95f5ece3c250f0b73c5a8dcab8dc208bd3ad15d41bcfb73fc4ed8ed7d0be5d
3
+ size 369241872
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/pretrained_model/train_config.json ADDED
@@ -0,0 +1,217 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "shylee/so100_pengripG",
4
+ "root": "/SSD/LSY/lerobot_dataset/pengripG",
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
+ }
62
+ },
63
+ "revision": null,
64
+ "use_imagenet_stats": true,
65
+ "video_backend": "torchcodec"
66
+ },
67
+ "env": null,
68
+ "policy": {
69
+ "type": "diffusion",
70
+ "n_obs_steps": 2,
71
+ "normalization_mapping": {
72
+ "VISUAL": "MEAN_STD",
73
+ "STATE": "MIN_MAX",
74
+ "ACTION": "MIN_MAX"
75
+ },
76
+ "input_features": {
77
+ "observation.state": {
78
+ "type": "STATE",
79
+ "shape": [
80
+ 6
81
+ ]
82
+ },
83
+ "observation.images.FrontCam": {
84
+ "type": "VISUAL",
85
+ "shape": [
86
+ 3,
87
+ 480,
88
+ 640
89
+ ]
90
+ },
91
+ "observation.images.TopCam": {
92
+ "type": "VISUAL",
93
+ "shape": [
94
+ 3,
95
+ 480,
96
+ 640
97
+ ]
98
+ },
99
+ "observation.images.WristCam": {
100
+ "type": "VISUAL",
101
+ "shape": [
102
+ 3,
103
+ 480,
104
+ 640
105
+ ]
106
+ }
107
+ },
108
+ "output_features": {
109
+ "action": {
110
+ "type": "ACTION",
111
+ "shape": [
112
+ 6
113
+ ]
114
+ }
115
+ },
116
+ "device": "cuda",
117
+ "use_amp": false,
118
+ "horizon": 16,
119
+ "n_action_steps": 8,
120
+ "drop_n_last_frames": 7,
121
+ "vision_backbone": "resnet50",
122
+ "crop_shape": [
123
+ 224,
124
+ 224
125
+ ],
126
+ "crop_is_random": false,
127
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
128
+ "use_group_norm": false,
129
+ "spatial_softmax_num_keypoints": 32,
130
+ "use_separate_rgb_encoder_per_camera": false,
131
+ "down_dims": [
132
+ 256,
133
+ 512,
134
+ 1024
135
+ ],
136
+ "kernel_size": 5,
137
+ "n_groups": 8,
138
+ "diffusion_step_embed_dim": 128,
139
+ "use_film_scale_modulation": true,
140
+ "noise_scheduler_type": "DDIM",
141
+ "num_train_timesteps": 100,
142
+ "beta_schedule": "squaredcos_cap_v2",
143
+ "beta_start": 0.0001,
144
+ "beta_end": 0.02,
145
+ "prediction_type": "epsilon",
146
+ "clip_sample": true,
147
+ "clip_sample_range": 1.0,
148
+ "num_inference_steps": 10,
149
+ "do_mask_loss_for_padding": false,
150
+ "optimizer_lr": 0.0001,
151
+ "optimizer_betas": [
152
+ 0.95,
153
+ 0.999
154
+ ],
155
+ "optimizer_eps": 1e-08,
156
+ "optimizer_weight_decay": 1e-06,
157
+ "scheduler_name": "cosine",
158
+ "scheduler_warmup_steps": 500,
159
+ "pre_resize_shape": [
160
+ 224,
161
+ 224
162
+ ],
163
+ "freeze_vision_backbone": false,
164
+ "use_ac": false,
165
+ "ac_threshold": 0.99,
166
+ "use_pcg": false,
167
+ "pcg_weight": 0.1,
168
+ "use_bid": false,
169
+ "bid_batch_size": 15,
170
+ "bid_mode_size": 3,
171
+ "bid_temporal_coherence": 0.5,
172
+ "weak_model_path": null,
173
+ "weak_model_config": null
174
+ },
175
+ "output_dir": "/SSD/LSY/lerobot_model/DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
176
+ "job_name": "DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
177
+ "resume": true,
178
+ "seed": 1000,
179
+ "num_workers": 4,
180
+ "batch_size": 16,
181
+ "steps": 900000,
182
+ "eval_freq": 20000,
183
+ "log_freq": 200,
184
+ "save_checkpoint": true,
185
+ "save_freq": 20000,
186
+ "use_policy_training_preset": true,
187
+ "optimizer": {
188
+ "type": "adam",
189
+ "lr": 0.0001,
190
+ "weight_decay": 1e-06,
191
+ "grad_clip_norm": 10.0,
192
+ "betas": [
193
+ 0.95,
194
+ 0.999
195
+ ],
196
+ "eps": 1e-08
197
+ },
198
+ "scheduler": {
199
+ "type": "diffuser",
200
+ "num_warmup_steps": 500,
201
+ "name": "cosine"
202
+ },
203
+ "eval": {
204
+ "n_episodes": 50,
205
+ "batch_size": 50,
206
+ "use_async_envs": false
207
+ },
208
+ "wandb": {
209
+ "enable": true,
210
+ "disable_artifact": false,
211
+ "project": "lerobot",
212
+ "entity": null,
213
+ "notes": null,
214
+ "run_id": "bqr8y636",
215
+ "mode": null
216
+ }
217
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/training_state/optimizer_param_groups.json ADDED
@@ -0,0 +1,331 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "lr": 7.800949179930113e-05,
4
+ "betas": [
5
+ 0.95,
6
+ 0.999
7
+ ],
8
+ "eps": 1e-08,
9
+ "weight_decay": 1e-06,
10
+ "amsgrad": false,
11
+ "maximize": false,
12
+ "foreach": null,
13
+ "capturable": false,
14
+ "differentiable": false,
15
+ "fused": null,
16
+ "initial_lr": 0.0001,
17
+ "params": [
18
+ 0,
19
+ 1,
20
+ 2,
21
+ 3,
22
+ 4,
23
+ 5,
24
+ 6,
25
+ 7,
26
+ 8,
27
+ 9,
28
+ 10,
29
+ 11,
30
+ 12,
31
+ 13,
32
+ 14,
33
+ 15,
34
+ 16,
35
+ 17,
36
+ 18,
37
+ 19,
38
+ 20,
39
+ 21,
40
+ 22,
41
+ 23,
42
+ 24,
43
+ 25,
44
+ 26,
45
+ 27,
46
+ 28,
47
+ 29,
48
+ 30,
49
+ 31,
50
+ 32,
51
+ 33,
52
+ 34,
53
+ 35,
54
+ 36,
55
+ 37,
56
+ 38,
57
+ 39,
58
+ 40,
59
+ 41,
60
+ 42,
61
+ 43,
62
+ 44,
63
+ 45,
64
+ 46,
65
+ 47,
66
+ 48,
67
+ 49,
68
+ 50,
69
+ 51,
70
+ 52,
71
+ 53,
72
+ 54,
73
+ 55,
74
+ 56,
75
+ 57,
76
+ 58,
77
+ 59,
78
+ 60,
79
+ 61,
80
+ 62,
81
+ 63,
82
+ 64,
83
+ 65,
84
+ 66,
85
+ 67,
86
+ 68,
87
+ 69,
88
+ 70,
89
+ 71,
90
+ 72,
91
+ 73,
92
+ 74,
93
+ 75,
94
+ 76,
95
+ 77,
96
+ 78,
97
+ 79,
98
+ 80,
99
+ 81,
100
+ 82,
101
+ 83,
102
+ 84,
103
+ 85,
104
+ 86,
105
+ 87,
106
+ 88,
107
+ 89,
108
+ 90,
109
+ 91,
110
+ 92,
111
+ 93,
112
+ 94,
113
+ 95,
114
+ 96,
115
+ 97,
116
+ 98,
117
+ 99,
118
+ 100,
119
+ 101,
120
+ 102,
121
+ 103,
122
+ 104,
123
+ 105,
124
+ 106,
125
+ 107,
126
+ 108,
127
+ 109,
128
+ 110,
129
+ 111,
130
+ 112,
131
+ 113,
132
+ 114,
133
+ 115,
134
+ 116,
135
+ 117,
136
+ 118,
137
+ 119,
138
+ 120,
139
+ 121,
140
+ 122,
141
+ 123,
142
+ 124,
143
+ 125,
144
+ 126,
145
+ 127,
146
+ 128,
147
+ 129,
148
+ 130,
149
+ 131,
150
+ 132,
151
+ 133,
152
+ 134,
153
+ 135,
154
+ 136,
155
+ 137,
156
+ 138,
157
+ 139,
158
+ 140,
159
+ 141,
160
+ 142,
161
+ 143,
162
+ 144,
163
+ 145,
164
+ 146,
165
+ 147,
166
+ 148,
167
+ 149,
168
+ 150,
169
+ 151,
170
+ 152,
171
+ 153,
172
+ 154,
173
+ 155,
174
+ 156,
175
+ 157,
176
+ 158,
177
+ 159,
178
+ 160,
179
+ 161,
180
+ 162,
181
+ 163,
182
+ 164,
183
+ 165,
184
+ 166,
185
+ 167,
186
+ 168,
187
+ 169,
188
+ 170,
189
+ 171,
190
+ 172,
191
+ 173,
192
+ 174,
193
+ 175,
194
+ 176,
195
+ 177,
196
+ 178,
197
+ 179,
198
+ 180,
199
+ 181,
200
+ 182,
201
+ 183,
202
+ 184,
203
+ 185,
204
+ 186,
205
+ 187,
206
+ 188,
207
+ 189,
208
+ 190,
209
+ 191,
210
+ 192,
211
+ 193,
212
+ 194,
213
+ 195,
214
+ 196,
215
+ 197,
216
+ 198,
217
+ 199,
218
+ 200,
219
+ 201,
220
+ 202,
221
+ 203,
222
+ 204,
223
+ 205,
224
+ 206,
225
+ 207,
226
+ 208,
227
+ 209,
228
+ 210,
229
+ 211,
230
+ 212,
231
+ 213,
232
+ 214,
233
+ 215,
234
+ 216,
235
+ 217,
236
+ 218,
237
+ 219,
238
+ 220,
239
+ 221,
240
+ 222,
241
+ 223,
242
+ 224,
243
+ 225,
244
+ 226,
245
+ 227,
246
+ 228,
247
+ 229,
248
+ 230,
249
+ 231,
250
+ 232,
251
+ 233,
252
+ 234,
253
+ 235,
254
+ 236,
255
+ 237,
256
+ 238,
257
+ 239,
258
+ 240,
259
+ 241,
260
+ 242,
261
+ 243,
262
+ 244,
263
+ 245,
264
+ 246,
265
+ 247,
266
+ 248,
267
+ 249,
268
+ 250,
269
+ 251,
270
+ 252,
271
+ 253,
272
+ 254,
273
+ 255,
274
+ 256,
275
+ 257,
276
+ 258,
277
+ 259,
278
+ 260,
279
+ 261,
280
+ 262,
281
+ 263,
282
+ 264,
283
+ 265,
284
+ 266,
285
+ 267,
286
+ 268,
287
+ 269,
288
+ 270,
289
+ 271,
290
+ 272,
291
+ 273,
292
+ 274,
293
+ 275,
294
+ 276,
295
+ 277,
296
+ 278,
297
+ 279,
298
+ 280,
299
+ 281,
300
+ 282,
301
+ 283,
302
+ 284,
303
+ 285,
304
+ 286,
305
+ 287,
306
+ 288,
307
+ 289,
308
+ 290,
309
+ 291,
310
+ 292,
311
+ 293,
312
+ 294,
313
+ 295,
314
+ 296,
315
+ 297,
316
+ 298,
317
+ 299,
318
+ 300,
319
+ 301,
320
+ 302,
321
+ 303,
322
+ 304,
323
+ 305,
324
+ 306,
325
+ 307,
326
+ 308,
327
+ 309,
328
+ 310
329
+ ]
330
+ }
331
+ ]
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/training_state/optimizer_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08263ee32e78dc147512ed8e06a4dcc7f513da5e638c7f312988b069ebf2c964
3
+ size 738026076
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/training_state/rng_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30f5dcf9d86ab6d40c5f4424b4efb3d8522ef08cfa4ba99e1b18727deff4bc7a
3
+ size 15708
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/training_state/scheduler_state.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_lrs": [
3
+ 0.0001
4
+ ],
5
+ "last_epoch": 280000,
6
+ "verbose": false,
7
+ "_step_count": 280001,
8
+ "_get_lr_called_within_step": false,
9
+ "_last_lr": [
10
+ 7.800949179930113e-05
11
+ ],
12
+ "lr_lambdas": [
13
+ null
14
+ ]
15
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/280000/training_state/training_step.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "step": 280000
3
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/pretrained_model/config.json ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "diffusion",
3
+ "n_obs_steps": 2,
4
+ "normalization_mapping": {
5
+ "VISUAL": "MEAN_STD",
6
+ "STATE": "MIN_MAX",
7
+ "ACTION": "MIN_MAX"
8
+ },
9
+ "input_features": {
10
+ "observation.state": {
11
+ "type": "STATE",
12
+ "shape": [
13
+ 6
14
+ ]
15
+ },
16
+ "observation.images.FrontCam": {
17
+ "type": "VISUAL",
18
+ "shape": [
19
+ 3,
20
+ 480,
21
+ 640
22
+ ]
23
+ },
24
+ "observation.images.TopCam": {
25
+ "type": "VISUAL",
26
+ "shape": [
27
+ 3,
28
+ 480,
29
+ 640
30
+ ]
31
+ },
32
+ "observation.images.WristCam": {
33
+ "type": "VISUAL",
34
+ "shape": [
35
+ 3,
36
+ 480,
37
+ 640
38
+ ]
39
+ }
40
+ },
41
+ "output_features": {
42
+ "action": {
43
+ "type": "ACTION",
44
+ "shape": [
45
+ 6
46
+ ]
47
+ }
48
+ },
49
+ "device": "cuda",
50
+ "use_amp": false,
51
+ "horizon": 16,
52
+ "n_action_steps": 8,
53
+ "drop_n_last_frames": 7,
54
+ "vision_backbone": "resnet50",
55
+ "crop_shape": [
56
+ 224,
57
+ 224
58
+ ],
59
+ "crop_is_random": false,
60
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
61
+ "use_group_norm": false,
62
+ "spatial_softmax_num_keypoints": 32,
63
+ "use_separate_rgb_encoder_per_camera": false,
64
+ "down_dims": [
65
+ 256,
66
+ 512,
67
+ 1024
68
+ ],
69
+ "kernel_size": 5,
70
+ "n_groups": 8,
71
+ "diffusion_step_embed_dim": 128,
72
+ "use_film_scale_modulation": true,
73
+ "noise_scheduler_type": "DDIM",
74
+ "num_train_timesteps": 100,
75
+ "beta_schedule": "squaredcos_cap_v2",
76
+ "beta_start": 0.0001,
77
+ "beta_end": 0.02,
78
+ "prediction_type": "epsilon",
79
+ "clip_sample": true,
80
+ "clip_sample_range": 1.0,
81
+ "num_inference_steps": 10,
82
+ "do_mask_loss_for_padding": false,
83
+ "optimizer_lr": 0.0001,
84
+ "optimizer_betas": [
85
+ 0.95,
86
+ 0.999
87
+ ],
88
+ "optimizer_eps": 1e-08,
89
+ "optimizer_weight_decay": 1e-06,
90
+ "scheduler_name": "cosine",
91
+ "scheduler_warmup_steps": 500,
92
+ "pre_resize_shape": [
93
+ 224,
94
+ 224
95
+ ],
96
+ "freeze_vision_backbone": false,
97
+ "use_ac": false,
98
+ "ac_threshold": 0.99,
99
+ "use_pcg": false,
100
+ "pcg_weight": 0.1,
101
+ "use_bid": false,
102
+ "bid_batch_size": 15,
103
+ "bid_mode_size": 3,
104
+ "bid_temporal_coherence": 0.5,
105
+ "weak_model_path": null,
106
+ "weak_model_config": null
107
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/pretrained_model/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a36b8b3c5a37c480033a8585ce7c60a66efd7ecf359ea65d993d8ea5e843194
3
+ size 369241872
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/pretrained_model/train_config.json ADDED
@@ -0,0 +1,217 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "shylee/so100_pengripG",
4
+ "root": "/SSD/LSY/lerobot_dataset/pengripG",
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
+ }
62
+ },
63
+ "revision": null,
64
+ "use_imagenet_stats": true,
65
+ "video_backend": "torchcodec"
66
+ },
67
+ "env": null,
68
+ "policy": {
69
+ "type": "diffusion",
70
+ "n_obs_steps": 2,
71
+ "normalization_mapping": {
72
+ "VISUAL": "MEAN_STD",
73
+ "STATE": "MIN_MAX",
74
+ "ACTION": "MIN_MAX"
75
+ },
76
+ "input_features": {
77
+ "observation.state": {
78
+ "type": "STATE",
79
+ "shape": [
80
+ 6
81
+ ]
82
+ },
83
+ "observation.images.FrontCam": {
84
+ "type": "VISUAL",
85
+ "shape": [
86
+ 3,
87
+ 480,
88
+ 640
89
+ ]
90
+ },
91
+ "observation.images.TopCam": {
92
+ "type": "VISUAL",
93
+ "shape": [
94
+ 3,
95
+ 480,
96
+ 640
97
+ ]
98
+ },
99
+ "observation.images.WristCam": {
100
+ "type": "VISUAL",
101
+ "shape": [
102
+ 3,
103
+ 480,
104
+ 640
105
+ ]
106
+ }
107
+ },
108
+ "output_features": {
109
+ "action": {
110
+ "type": "ACTION",
111
+ "shape": [
112
+ 6
113
+ ]
114
+ }
115
+ },
116
+ "device": "cuda",
117
+ "use_amp": false,
118
+ "horizon": 16,
119
+ "n_action_steps": 8,
120
+ "drop_n_last_frames": 7,
121
+ "vision_backbone": "resnet50",
122
+ "crop_shape": [
123
+ 224,
124
+ 224
125
+ ],
126
+ "crop_is_random": false,
127
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
128
+ "use_group_norm": false,
129
+ "spatial_softmax_num_keypoints": 32,
130
+ "use_separate_rgb_encoder_per_camera": false,
131
+ "down_dims": [
132
+ 256,
133
+ 512,
134
+ 1024
135
+ ],
136
+ "kernel_size": 5,
137
+ "n_groups": 8,
138
+ "diffusion_step_embed_dim": 128,
139
+ "use_film_scale_modulation": true,
140
+ "noise_scheduler_type": "DDIM",
141
+ "num_train_timesteps": 100,
142
+ "beta_schedule": "squaredcos_cap_v2",
143
+ "beta_start": 0.0001,
144
+ "beta_end": 0.02,
145
+ "prediction_type": "epsilon",
146
+ "clip_sample": true,
147
+ "clip_sample_range": 1.0,
148
+ "num_inference_steps": 10,
149
+ "do_mask_loss_for_padding": false,
150
+ "optimizer_lr": 0.0001,
151
+ "optimizer_betas": [
152
+ 0.95,
153
+ 0.999
154
+ ],
155
+ "optimizer_eps": 1e-08,
156
+ "optimizer_weight_decay": 1e-06,
157
+ "scheduler_name": "cosine",
158
+ "scheduler_warmup_steps": 500,
159
+ "pre_resize_shape": [
160
+ 224,
161
+ 224
162
+ ],
163
+ "freeze_vision_backbone": false,
164
+ "use_ac": false,
165
+ "ac_threshold": 0.99,
166
+ "use_pcg": false,
167
+ "pcg_weight": 0.1,
168
+ "use_bid": false,
169
+ "bid_batch_size": 15,
170
+ "bid_mode_size": 3,
171
+ "bid_temporal_coherence": 0.5,
172
+ "weak_model_path": null,
173
+ "weak_model_config": null
174
+ },
175
+ "output_dir": "/SSD/LSY/lerobot_model/DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
176
+ "job_name": "DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
177
+ "resume": true,
178
+ "seed": 1000,
179
+ "num_workers": 4,
180
+ "batch_size": 16,
181
+ "steps": 900000,
182
+ "eval_freq": 20000,
183
+ "log_freq": 200,
184
+ "save_checkpoint": true,
185
+ "save_freq": 20000,
186
+ "use_policy_training_preset": true,
187
+ "optimizer": {
188
+ "type": "adam",
189
+ "lr": 0.0001,
190
+ "weight_decay": 1e-06,
191
+ "grad_clip_norm": 10.0,
192
+ "betas": [
193
+ 0.95,
194
+ 0.999
195
+ ],
196
+ "eps": 1e-08
197
+ },
198
+ "scheduler": {
199
+ "type": "diffuser",
200
+ "num_warmup_steps": 500,
201
+ "name": "cosine"
202
+ },
203
+ "eval": {
204
+ "n_episodes": 50,
205
+ "batch_size": 50,
206
+ "use_async_envs": false
207
+ },
208
+ "wandb": {
209
+ "enable": true,
210
+ "disable_artifact": false,
211
+ "project": "lerobot",
212
+ "entity": null,
213
+ "notes": null,
214
+ "run_id": "bqr8y636",
215
+ "mode": null
216
+ }
217
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/training_state/optimizer_param_groups.json ADDED
@@ -0,0 +1,331 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "lr": 7.505039436856827e-05,
4
+ "betas": [
5
+ 0.95,
6
+ 0.999
7
+ ],
8
+ "eps": 1e-08,
9
+ "weight_decay": 1e-06,
10
+ "amsgrad": false,
11
+ "maximize": false,
12
+ "foreach": null,
13
+ "capturable": false,
14
+ "differentiable": false,
15
+ "fused": null,
16
+ "initial_lr": 0.0001,
17
+ "params": [
18
+ 0,
19
+ 1,
20
+ 2,
21
+ 3,
22
+ 4,
23
+ 5,
24
+ 6,
25
+ 7,
26
+ 8,
27
+ 9,
28
+ 10,
29
+ 11,
30
+ 12,
31
+ 13,
32
+ 14,
33
+ 15,
34
+ 16,
35
+ 17,
36
+ 18,
37
+ 19,
38
+ 20,
39
+ 21,
40
+ 22,
41
+ 23,
42
+ 24,
43
+ 25,
44
+ 26,
45
+ 27,
46
+ 28,
47
+ 29,
48
+ 30,
49
+ 31,
50
+ 32,
51
+ 33,
52
+ 34,
53
+ 35,
54
+ 36,
55
+ 37,
56
+ 38,
57
+ 39,
58
+ 40,
59
+ 41,
60
+ 42,
61
+ 43,
62
+ 44,
63
+ 45,
64
+ 46,
65
+ 47,
66
+ 48,
67
+ 49,
68
+ 50,
69
+ 51,
70
+ 52,
71
+ 53,
72
+ 54,
73
+ 55,
74
+ 56,
75
+ 57,
76
+ 58,
77
+ 59,
78
+ 60,
79
+ 61,
80
+ 62,
81
+ 63,
82
+ 64,
83
+ 65,
84
+ 66,
85
+ 67,
86
+ 68,
87
+ 69,
88
+ 70,
89
+ 71,
90
+ 72,
91
+ 73,
92
+ 74,
93
+ 75,
94
+ 76,
95
+ 77,
96
+ 78,
97
+ 79,
98
+ 80,
99
+ 81,
100
+ 82,
101
+ 83,
102
+ 84,
103
+ 85,
104
+ 86,
105
+ 87,
106
+ 88,
107
+ 89,
108
+ 90,
109
+ 91,
110
+ 92,
111
+ 93,
112
+ 94,
113
+ 95,
114
+ 96,
115
+ 97,
116
+ 98,
117
+ 99,
118
+ 100,
119
+ 101,
120
+ 102,
121
+ 103,
122
+ 104,
123
+ 105,
124
+ 106,
125
+ 107,
126
+ 108,
127
+ 109,
128
+ 110,
129
+ 111,
130
+ 112,
131
+ 113,
132
+ 114,
133
+ 115,
134
+ 116,
135
+ 117,
136
+ 118,
137
+ 119,
138
+ 120,
139
+ 121,
140
+ 122,
141
+ 123,
142
+ 124,
143
+ 125,
144
+ 126,
145
+ 127,
146
+ 128,
147
+ 129,
148
+ 130,
149
+ 131,
150
+ 132,
151
+ 133,
152
+ 134,
153
+ 135,
154
+ 136,
155
+ 137,
156
+ 138,
157
+ 139,
158
+ 140,
159
+ 141,
160
+ 142,
161
+ 143,
162
+ 144,
163
+ 145,
164
+ 146,
165
+ 147,
166
+ 148,
167
+ 149,
168
+ 150,
169
+ 151,
170
+ 152,
171
+ 153,
172
+ 154,
173
+ 155,
174
+ 156,
175
+ 157,
176
+ 158,
177
+ 159,
178
+ 160,
179
+ 161,
180
+ 162,
181
+ 163,
182
+ 164,
183
+ 165,
184
+ 166,
185
+ 167,
186
+ 168,
187
+ 169,
188
+ 170,
189
+ 171,
190
+ 172,
191
+ 173,
192
+ 174,
193
+ 175,
194
+ 176,
195
+ 177,
196
+ 178,
197
+ 179,
198
+ 180,
199
+ 181,
200
+ 182,
201
+ 183,
202
+ 184,
203
+ 185,
204
+ 186,
205
+ 187,
206
+ 188,
207
+ 189,
208
+ 190,
209
+ 191,
210
+ 192,
211
+ 193,
212
+ 194,
213
+ 195,
214
+ 196,
215
+ 197,
216
+ 198,
217
+ 199,
218
+ 200,
219
+ 201,
220
+ 202,
221
+ 203,
222
+ 204,
223
+ 205,
224
+ 206,
225
+ 207,
226
+ 208,
227
+ 209,
228
+ 210,
229
+ 211,
230
+ 212,
231
+ 213,
232
+ 214,
233
+ 215,
234
+ 216,
235
+ 217,
236
+ 218,
237
+ 219,
238
+ 220,
239
+ 221,
240
+ 222,
241
+ 223,
242
+ 224,
243
+ 225,
244
+ 226,
245
+ 227,
246
+ 228,
247
+ 229,
248
+ 230,
249
+ 231,
250
+ 232,
251
+ 233,
252
+ 234,
253
+ 235,
254
+ 236,
255
+ 237,
256
+ 238,
257
+ 239,
258
+ 240,
259
+ 241,
260
+ 242,
261
+ 243,
262
+ 244,
263
+ 245,
264
+ 246,
265
+ 247,
266
+ 248,
267
+ 249,
268
+ 250,
269
+ 251,
270
+ 252,
271
+ 253,
272
+ 254,
273
+ 255,
274
+ 256,
275
+ 257,
276
+ 258,
277
+ 259,
278
+ 260,
279
+ 261,
280
+ 262,
281
+ 263,
282
+ 264,
283
+ 265,
284
+ 266,
285
+ 267,
286
+ 268,
287
+ 269,
288
+ 270,
289
+ 271,
290
+ 272,
291
+ 273,
292
+ 274,
293
+ 275,
294
+ 276,
295
+ 277,
296
+ 278,
297
+ 279,
298
+ 280,
299
+ 281,
300
+ 282,
301
+ 283,
302
+ 284,
303
+ 285,
304
+ 286,
305
+ 287,
306
+ 288,
307
+ 289,
308
+ 290,
309
+ 291,
310
+ 292,
311
+ 293,
312
+ 294,
313
+ 295,
314
+ 296,
315
+ 297,
316
+ 298,
317
+ 299,
318
+ 300,
319
+ 301,
320
+ 302,
321
+ 303,
322
+ 304,
323
+ 305,
324
+ 306,
325
+ 307,
326
+ 308,
327
+ 309,
328
+ 310
329
+ ]
330
+ }
331
+ ]
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/training_state/optimizer_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52a34fb6c5f0bfa2464b5e12e753f5baaa438789033b7fe30bf7e5287c7413b8
3
+ size 738026076
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/training_state/rng_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d70ec757ce50e217e55e9f9af4e6bbc64f6359f812489fc68b023607554fc85
3
+ size 15708
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/training_state/scheduler_state.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_lrs": [
3
+ 0.0001
4
+ ],
5
+ "last_epoch": 300000,
6
+ "verbose": false,
7
+ "_step_count": 300001,
8
+ "_get_lr_called_within_step": false,
9
+ "_last_lr": [
10
+ 7.505039436856827e-05
11
+ ],
12
+ "lr_lambdas": [
13
+ null
14
+ ]
15
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/300000/training_state/training_step.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "step": 300000
3
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/pretrained_model/config.json ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "diffusion",
3
+ "n_obs_steps": 2,
4
+ "normalization_mapping": {
5
+ "VISUAL": "MEAN_STD",
6
+ "STATE": "MIN_MAX",
7
+ "ACTION": "MIN_MAX"
8
+ },
9
+ "input_features": {
10
+ "observation.state": {
11
+ "type": "STATE",
12
+ "shape": [
13
+ 6
14
+ ]
15
+ },
16
+ "observation.images.FrontCam": {
17
+ "type": "VISUAL",
18
+ "shape": [
19
+ 3,
20
+ 480,
21
+ 640
22
+ ]
23
+ },
24
+ "observation.images.TopCam": {
25
+ "type": "VISUAL",
26
+ "shape": [
27
+ 3,
28
+ 480,
29
+ 640
30
+ ]
31
+ },
32
+ "observation.images.WristCam": {
33
+ "type": "VISUAL",
34
+ "shape": [
35
+ 3,
36
+ 480,
37
+ 640
38
+ ]
39
+ }
40
+ },
41
+ "output_features": {
42
+ "action": {
43
+ "type": "ACTION",
44
+ "shape": [
45
+ 6
46
+ ]
47
+ }
48
+ },
49
+ "device": "cuda",
50
+ "use_amp": false,
51
+ "horizon": 16,
52
+ "n_action_steps": 8,
53
+ "drop_n_last_frames": 7,
54
+ "vision_backbone": "resnet50",
55
+ "crop_shape": [
56
+ 224,
57
+ 224
58
+ ],
59
+ "crop_is_random": false,
60
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
61
+ "use_group_norm": false,
62
+ "spatial_softmax_num_keypoints": 32,
63
+ "use_separate_rgb_encoder_per_camera": false,
64
+ "down_dims": [
65
+ 256,
66
+ 512,
67
+ 1024
68
+ ],
69
+ "kernel_size": 5,
70
+ "n_groups": 8,
71
+ "diffusion_step_embed_dim": 128,
72
+ "use_film_scale_modulation": true,
73
+ "noise_scheduler_type": "DDIM",
74
+ "num_train_timesteps": 100,
75
+ "beta_schedule": "squaredcos_cap_v2",
76
+ "beta_start": 0.0001,
77
+ "beta_end": 0.02,
78
+ "prediction_type": "epsilon",
79
+ "clip_sample": true,
80
+ "clip_sample_range": 1.0,
81
+ "num_inference_steps": 10,
82
+ "do_mask_loss_for_padding": false,
83
+ "optimizer_lr": 0.0001,
84
+ "optimizer_betas": [
85
+ 0.95,
86
+ 0.999
87
+ ],
88
+ "optimizer_eps": 1e-08,
89
+ "optimizer_weight_decay": 1e-06,
90
+ "scheduler_name": "cosine",
91
+ "scheduler_warmup_steps": 500,
92
+ "pre_resize_shape": [
93
+ 224,
94
+ 224
95
+ ],
96
+ "freeze_vision_backbone": false,
97
+ "use_ac": false,
98
+ "ac_threshold": 0.99,
99
+ "use_pcg": false,
100
+ "pcg_weight": 0.1,
101
+ "use_bid": false,
102
+ "bid_batch_size": 15,
103
+ "bid_mode_size": 3,
104
+ "bid_temporal_coherence": 0.5,
105
+ "weak_model_path": null,
106
+ "weak_model_config": null
107
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/pretrained_model/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27af29a48fd4860aeab76e70a002fe27fa2f168929941fc46eb300a0a6a55970
3
+ size 369241872
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/pretrained_model/train_config.json ADDED
@@ -0,0 +1,217 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "shylee/so100_pengripG",
4
+ "root": "/SSD/LSY/lerobot_dataset/pengripG",
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
+ }
62
+ },
63
+ "revision": null,
64
+ "use_imagenet_stats": true,
65
+ "video_backend": "torchcodec"
66
+ },
67
+ "env": null,
68
+ "policy": {
69
+ "type": "diffusion",
70
+ "n_obs_steps": 2,
71
+ "normalization_mapping": {
72
+ "VISUAL": "MEAN_STD",
73
+ "STATE": "MIN_MAX",
74
+ "ACTION": "MIN_MAX"
75
+ },
76
+ "input_features": {
77
+ "observation.state": {
78
+ "type": "STATE",
79
+ "shape": [
80
+ 6
81
+ ]
82
+ },
83
+ "observation.images.FrontCam": {
84
+ "type": "VISUAL",
85
+ "shape": [
86
+ 3,
87
+ 480,
88
+ 640
89
+ ]
90
+ },
91
+ "observation.images.TopCam": {
92
+ "type": "VISUAL",
93
+ "shape": [
94
+ 3,
95
+ 480,
96
+ 640
97
+ ]
98
+ },
99
+ "observation.images.WristCam": {
100
+ "type": "VISUAL",
101
+ "shape": [
102
+ 3,
103
+ 480,
104
+ 640
105
+ ]
106
+ }
107
+ },
108
+ "output_features": {
109
+ "action": {
110
+ "type": "ACTION",
111
+ "shape": [
112
+ 6
113
+ ]
114
+ }
115
+ },
116
+ "device": "cuda",
117
+ "use_amp": false,
118
+ "horizon": 16,
119
+ "n_action_steps": 8,
120
+ "drop_n_last_frames": 7,
121
+ "vision_backbone": "resnet50",
122
+ "crop_shape": [
123
+ 224,
124
+ 224
125
+ ],
126
+ "crop_is_random": false,
127
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
128
+ "use_group_norm": false,
129
+ "spatial_softmax_num_keypoints": 32,
130
+ "use_separate_rgb_encoder_per_camera": false,
131
+ "down_dims": [
132
+ 256,
133
+ 512,
134
+ 1024
135
+ ],
136
+ "kernel_size": 5,
137
+ "n_groups": 8,
138
+ "diffusion_step_embed_dim": 128,
139
+ "use_film_scale_modulation": true,
140
+ "noise_scheduler_type": "DDIM",
141
+ "num_train_timesteps": 100,
142
+ "beta_schedule": "squaredcos_cap_v2",
143
+ "beta_start": 0.0001,
144
+ "beta_end": 0.02,
145
+ "prediction_type": "epsilon",
146
+ "clip_sample": true,
147
+ "clip_sample_range": 1.0,
148
+ "num_inference_steps": 10,
149
+ "do_mask_loss_for_padding": false,
150
+ "optimizer_lr": 0.0001,
151
+ "optimizer_betas": [
152
+ 0.95,
153
+ 0.999
154
+ ],
155
+ "optimizer_eps": 1e-08,
156
+ "optimizer_weight_decay": 1e-06,
157
+ "scheduler_name": "cosine",
158
+ "scheduler_warmup_steps": 500,
159
+ "pre_resize_shape": [
160
+ 224,
161
+ 224
162
+ ],
163
+ "freeze_vision_backbone": false,
164
+ "use_ac": false,
165
+ "ac_threshold": 0.99,
166
+ "use_pcg": false,
167
+ "pcg_weight": 0.1,
168
+ "use_bid": false,
169
+ "bid_batch_size": 15,
170
+ "bid_mode_size": 3,
171
+ "bid_temporal_coherence": 0.5,
172
+ "weak_model_path": null,
173
+ "weak_model_config": null
174
+ },
175
+ "output_dir": "/SSD/LSY/lerobot_model/DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
176
+ "job_name": "DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
177
+ "resume": true,
178
+ "seed": 1000,
179
+ "num_workers": 4,
180
+ "batch_size": 16,
181
+ "steps": 900000,
182
+ "eval_freq": 20000,
183
+ "log_freq": 200,
184
+ "save_checkpoint": true,
185
+ "save_freq": 20000,
186
+ "use_policy_training_preset": true,
187
+ "optimizer": {
188
+ "type": "adam",
189
+ "lr": 0.0001,
190
+ "weight_decay": 1e-06,
191
+ "grad_clip_norm": 10.0,
192
+ "betas": [
193
+ 0.95,
194
+ 0.999
195
+ ],
196
+ "eps": 1e-08
197
+ },
198
+ "scheduler": {
199
+ "type": "diffuser",
200
+ "num_warmup_steps": 500,
201
+ "name": "cosine"
202
+ },
203
+ "eval": {
204
+ "n_episodes": 50,
205
+ "batch_size": 50,
206
+ "use_async_envs": false
207
+ },
208
+ "wandb": {
209
+ "enable": true,
210
+ "disable_artifact": false,
211
+ "project": "lerobot",
212
+ "entity": null,
213
+ "notes": null,
214
+ "run_id": "bqr8y636",
215
+ "mode": null
216
+ }
217
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/training_state/optimizer_param_groups.json ADDED
@@ -0,0 +1,331 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "lr": 7.196911827329292e-05,
4
+ "betas": [
5
+ 0.95,
6
+ 0.999
7
+ ],
8
+ "eps": 1e-08,
9
+ "weight_decay": 1e-06,
10
+ "amsgrad": false,
11
+ "maximize": false,
12
+ "foreach": null,
13
+ "capturable": false,
14
+ "differentiable": false,
15
+ "fused": null,
16
+ "initial_lr": 0.0001,
17
+ "params": [
18
+ 0,
19
+ 1,
20
+ 2,
21
+ 3,
22
+ 4,
23
+ 5,
24
+ 6,
25
+ 7,
26
+ 8,
27
+ 9,
28
+ 10,
29
+ 11,
30
+ 12,
31
+ 13,
32
+ 14,
33
+ 15,
34
+ 16,
35
+ 17,
36
+ 18,
37
+ 19,
38
+ 20,
39
+ 21,
40
+ 22,
41
+ 23,
42
+ 24,
43
+ 25,
44
+ 26,
45
+ 27,
46
+ 28,
47
+ 29,
48
+ 30,
49
+ 31,
50
+ 32,
51
+ 33,
52
+ 34,
53
+ 35,
54
+ 36,
55
+ 37,
56
+ 38,
57
+ 39,
58
+ 40,
59
+ 41,
60
+ 42,
61
+ 43,
62
+ 44,
63
+ 45,
64
+ 46,
65
+ 47,
66
+ 48,
67
+ 49,
68
+ 50,
69
+ 51,
70
+ 52,
71
+ 53,
72
+ 54,
73
+ 55,
74
+ 56,
75
+ 57,
76
+ 58,
77
+ 59,
78
+ 60,
79
+ 61,
80
+ 62,
81
+ 63,
82
+ 64,
83
+ 65,
84
+ 66,
85
+ 67,
86
+ 68,
87
+ 69,
88
+ 70,
89
+ 71,
90
+ 72,
91
+ 73,
92
+ 74,
93
+ 75,
94
+ 76,
95
+ 77,
96
+ 78,
97
+ 79,
98
+ 80,
99
+ 81,
100
+ 82,
101
+ 83,
102
+ 84,
103
+ 85,
104
+ 86,
105
+ 87,
106
+ 88,
107
+ 89,
108
+ 90,
109
+ 91,
110
+ 92,
111
+ 93,
112
+ 94,
113
+ 95,
114
+ 96,
115
+ 97,
116
+ 98,
117
+ 99,
118
+ 100,
119
+ 101,
120
+ 102,
121
+ 103,
122
+ 104,
123
+ 105,
124
+ 106,
125
+ 107,
126
+ 108,
127
+ 109,
128
+ 110,
129
+ 111,
130
+ 112,
131
+ 113,
132
+ 114,
133
+ 115,
134
+ 116,
135
+ 117,
136
+ 118,
137
+ 119,
138
+ 120,
139
+ 121,
140
+ 122,
141
+ 123,
142
+ 124,
143
+ 125,
144
+ 126,
145
+ 127,
146
+ 128,
147
+ 129,
148
+ 130,
149
+ 131,
150
+ 132,
151
+ 133,
152
+ 134,
153
+ 135,
154
+ 136,
155
+ 137,
156
+ 138,
157
+ 139,
158
+ 140,
159
+ 141,
160
+ 142,
161
+ 143,
162
+ 144,
163
+ 145,
164
+ 146,
165
+ 147,
166
+ 148,
167
+ 149,
168
+ 150,
169
+ 151,
170
+ 152,
171
+ 153,
172
+ 154,
173
+ 155,
174
+ 156,
175
+ 157,
176
+ 158,
177
+ 159,
178
+ 160,
179
+ 161,
180
+ 162,
181
+ 163,
182
+ 164,
183
+ 165,
184
+ 166,
185
+ 167,
186
+ 168,
187
+ 169,
188
+ 170,
189
+ 171,
190
+ 172,
191
+ 173,
192
+ 174,
193
+ 175,
194
+ 176,
195
+ 177,
196
+ 178,
197
+ 179,
198
+ 180,
199
+ 181,
200
+ 182,
201
+ 183,
202
+ 184,
203
+ 185,
204
+ 186,
205
+ 187,
206
+ 188,
207
+ 189,
208
+ 190,
209
+ 191,
210
+ 192,
211
+ 193,
212
+ 194,
213
+ 195,
214
+ 196,
215
+ 197,
216
+ 198,
217
+ 199,
218
+ 200,
219
+ 201,
220
+ 202,
221
+ 203,
222
+ 204,
223
+ 205,
224
+ 206,
225
+ 207,
226
+ 208,
227
+ 209,
228
+ 210,
229
+ 211,
230
+ 212,
231
+ 213,
232
+ 214,
233
+ 215,
234
+ 216,
235
+ 217,
236
+ 218,
237
+ 219,
238
+ 220,
239
+ 221,
240
+ 222,
241
+ 223,
242
+ 224,
243
+ 225,
244
+ 226,
245
+ 227,
246
+ 228,
247
+ 229,
248
+ 230,
249
+ 231,
250
+ 232,
251
+ 233,
252
+ 234,
253
+ 235,
254
+ 236,
255
+ 237,
256
+ 238,
257
+ 239,
258
+ 240,
259
+ 241,
260
+ 242,
261
+ 243,
262
+ 244,
263
+ 245,
264
+ 246,
265
+ 247,
266
+ 248,
267
+ 249,
268
+ 250,
269
+ 251,
270
+ 252,
271
+ 253,
272
+ 254,
273
+ 255,
274
+ 256,
275
+ 257,
276
+ 258,
277
+ 259,
278
+ 260,
279
+ 261,
280
+ 262,
281
+ 263,
282
+ 264,
283
+ 265,
284
+ 266,
285
+ 267,
286
+ 268,
287
+ 269,
288
+ 270,
289
+ 271,
290
+ 272,
291
+ 273,
292
+ 274,
293
+ 275,
294
+ 276,
295
+ 277,
296
+ 278,
297
+ 279,
298
+ 280,
299
+ 281,
300
+ 282,
301
+ 283,
302
+ 284,
303
+ 285,
304
+ 286,
305
+ 287,
306
+ 288,
307
+ 289,
308
+ 290,
309
+ 291,
310
+ 292,
311
+ 293,
312
+ 294,
313
+ 295,
314
+ 296,
315
+ 297,
316
+ 298,
317
+ 299,
318
+ 300,
319
+ 301,
320
+ 302,
321
+ 303,
322
+ 304,
323
+ 305,
324
+ 306,
325
+ 307,
326
+ 308,
327
+ 309,
328
+ 310
329
+ ]
330
+ }
331
+ ]
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/training_state/optimizer_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c4e7660a3d8566e448c5a9866d9c04ce5000e3496e6dc7d88db46c6adbae472
3
+ size 738026076
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/training_state/rng_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:43e11a6c813c7229f1de118b102d59cdc29f4a502513ac858bd6e11d4b998a45
3
+ size 15708
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/training_state/scheduler_state.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_lrs": [
3
+ 0.0001
4
+ ],
5
+ "last_epoch": 320000,
6
+ "verbose": false,
7
+ "_step_count": 320001,
8
+ "_get_lr_called_within_step": false,
9
+ "_last_lr": [
10
+ 7.196911827329292e-05
11
+ ],
12
+ "lr_lambdas": [
13
+ null
14
+ ]
15
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/320000/training_state/training_step.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "step": 320000
3
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/pretrained_model/config.json ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "diffusion",
3
+ "n_obs_steps": 2,
4
+ "normalization_mapping": {
5
+ "VISUAL": "MEAN_STD",
6
+ "STATE": "MIN_MAX",
7
+ "ACTION": "MIN_MAX"
8
+ },
9
+ "input_features": {
10
+ "observation.state": {
11
+ "type": "STATE",
12
+ "shape": [
13
+ 6
14
+ ]
15
+ },
16
+ "observation.images.FrontCam": {
17
+ "type": "VISUAL",
18
+ "shape": [
19
+ 3,
20
+ 480,
21
+ 640
22
+ ]
23
+ },
24
+ "observation.images.TopCam": {
25
+ "type": "VISUAL",
26
+ "shape": [
27
+ 3,
28
+ 480,
29
+ 640
30
+ ]
31
+ },
32
+ "observation.images.WristCam": {
33
+ "type": "VISUAL",
34
+ "shape": [
35
+ 3,
36
+ 480,
37
+ 640
38
+ ]
39
+ }
40
+ },
41
+ "output_features": {
42
+ "action": {
43
+ "type": "ACTION",
44
+ "shape": [
45
+ 6
46
+ ]
47
+ }
48
+ },
49
+ "device": "cuda",
50
+ "use_amp": false,
51
+ "horizon": 16,
52
+ "n_action_steps": 8,
53
+ "drop_n_last_frames": 7,
54
+ "vision_backbone": "resnet50",
55
+ "crop_shape": [
56
+ 224,
57
+ 224
58
+ ],
59
+ "crop_is_random": false,
60
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
61
+ "use_group_norm": false,
62
+ "spatial_softmax_num_keypoints": 32,
63
+ "use_separate_rgb_encoder_per_camera": false,
64
+ "down_dims": [
65
+ 256,
66
+ 512,
67
+ 1024
68
+ ],
69
+ "kernel_size": 5,
70
+ "n_groups": 8,
71
+ "diffusion_step_embed_dim": 128,
72
+ "use_film_scale_modulation": true,
73
+ "noise_scheduler_type": "DDIM",
74
+ "num_train_timesteps": 100,
75
+ "beta_schedule": "squaredcos_cap_v2",
76
+ "beta_start": 0.0001,
77
+ "beta_end": 0.02,
78
+ "prediction_type": "epsilon",
79
+ "clip_sample": true,
80
+ "clip_sample_range": 1.0,
81
+ "num_inference_steps": 10,
82
+ "do_mask_loss_for_padding": false,
83
+ "optimizer_lr": 0.0001,
84
+ "optimizer_betas": [
85
+ 0.95,
86
+ 0.999
87
+ ],
88
+ "optimizer_eps": 1e-08,
89
+ "optimizer_weight_decay": 1e-06,
90
+ "scheduler_name": "cosine",
91
+ "scheduler_warmup_steps": 500,
92
+ "pre_resize_shape": [
93
+ 224,
94
+ 224
95
+ ],
96
+ "freeze_vision_backbone": false,
97
+ "use_ac": false,
98
+ "ac_threshold": 0.99,
99
+ "use_pcg": false,
100
+ "pcg_weight": 0.1,
101
+ "use_bid": false,
102
+ "bid_batch_size": 15,
103
+ "bid_mode_size": 3,
104
+ "bid_temporal_coherence": 0.5,
105
+ "weak_model_path": null,
106
+ "weak_model_config": null
107
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/pretrained_model/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:775f96b61fda929db323fa54b00410d267f901d0d53dfa7df896b64d661cd14d
3
+ size 369241872
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/pretrained_model/train_config.json ADDED
@@ -0,0 +1,217 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "shylee/so100_pengripG",
4
+ "root": "/SSD/LSY/lerobot_dataset/pengripG",
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
+ }
62
+ },
63
+ "revision": null,
64
+ "use_imagenet_stats": true,
65
+ "video_backend": "torchcodec"
66
+ },
67
+ "env": null,
68
+ "policy": {
69
+ "type": "diffusion",
70
+ "n_obs_steps": 2,
71
+ "normalization_mapping": {
72
+ "VISUAL": "MEAN_STD",
73
+ "STATE": "MIN_MAX",
74
+ "ACTION": "MIN_MAX"
75
+ },
76
+ "input_features": {
77
+ "observation.state": {
78
+ "type": "STATE",
79
+ "shape": [
80
+ 6
81
+ ]
82
+ },
83
+ "observation.images.FrontCam": {
84
+ "type": "VISUAL",
85
+ "shape": [
86
+ 3,
87
+ 480,
88
+ 640
89
+ ]
90
+ },
91
+ "observation.images.TopCam": {
92
+ "type": "VISUAL",
93
+ "shape": [
94
+ 3,
95
+ 480,
96
+ 640
97
+ ]
98
+ },
99
+ "observation.images.WristCam": {
100
+ "type": "VISUAL",
101
+ "shape": [
102
+ 3,
103
+ 480,
104
+ 640
105
+ ]
106
+ }
107
+ },
108
+ "output_features": {
109
+ "action": {
110
+ "type": "ACTION",
111
+ "shape": [
112
+ 6
113
+ ]
114
+ }
115
+ },
116
+ "device": "cuda",
117
+ "use_amp": false,
118
+ "horizon": 16,
119
+ "n_action_steps": 8,
120
+ "drop_n_last_frames": 7,
121
+ "vision_backbone": "resnet50",
122
+ "crop_shape": [
123
+ 224,
124
+ 224
125
+ ],
126
+ "crop_is_random": false,
127
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
128
+ "use_group_norm": false,
129
+ "spatial_softmax_num_keypoints": 32,
130
+ "use_separate_rgb_encoder_per_camera": false,
131
+ "down_dims": [
132
+ 256,
133
+ 512,
134
+ 1024
135
+ ],
136
+ "kernel_size": 5,
137
+ "n_groups": 8,
138
+ "diffusion_step_embed_dim": 128,
139
+ "use_film_scale_modulation": true,
140
+ "noise_scheduler_type": "DDIM",
141
+ "num_train_timesteps": 100,
142
+ "beta_schedule": "squaredcos_cap_v2",
143
+ "beta_start": 0.0001,
144
+ "beta_end": 0.02,
145
+ "prediction_type": "epsilon",
146
+ "clip_sample": true,
147
+ "clip_sample_range": 1.0,
148
+ "num_inference_steps": 10,
149
+ "do_mask_loss_for_padding": false,
150
+ "optimizer_lr": 0.0001,
151
+ "optimizer_betas": [
152
+ 0.95,
153
+ 0.999
154
+ ],
155
+ "optimizer_eps": 1e-08,
156
+ "optimizer_weight_decay": 1e-06,
157
+ "scheduler_name": "cosine",
158
+ "scheduler_warmup_steps": 500,
159
+ "pre_resize_shape": [
160
+ 224,
161
+ 224
162
+ ],
163
+ "freeze_vision_backbone": false,
164
+ "use_ac": false,
165
+ "ac_threshold": 0.99,
166
+ "use_pcg": false,
167
+ "pcg_weight": 0.1,
168
+ "use_bid": false,
169
+ "bid_batch_size": 15,
170
+ "bid_mode_size": 3,
171
+ "bid_temporal_coherence": 0.5,
172
+ "weak_model_path": null,
173
+ "weak_model_config": null
174
+ },
175
+ "output_dir": "/SSD/LSY/lerobot_model/DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
176
+ "job_name": "DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
177
+ "resume": true,
178
+ "seed": 1000,
179
+ "num_workers": 4,
180
+ "batch_size": 16,
181
+ "steps": 900000,
182
+ "eval_freq": 20000,
183
+ "log_freq": 200,
184
+ "save_checkpoint": true,
185
+ "save_freq": 20000,
186
+ "use_policy_training_preset": true,
187
+ "optimizer": {
188
+ "type": "adam",
189
+ "lr": 0.0001,
190
+ "weight_decay": 1e-06,
191
+ "grad_clip_norm": 10.0,
192
+ "betas": [
193
+ 0.95,
194
+ 0.999
195
+ ],
196
+ "eps": 1e-08
197
+ },
198
+ "scheduler": {
199
+ "type": "diffuser",
200
+ "num_warmup_steps": 500,
201
+ "name": "cosine"
202
+ },
203
+ "eval": {
204
+ "n_episodes": 50,
205
+ "batch_size": 50,
206
+ "use_async_envs": false
207
+ },
208
+ "wandb": {
209
+ "enable": true,
210
+ "disable_artifact": false,
211
+ "project": "lerobot",
212
+ "entity": null,
213
+ "notes": null,
214
+ "run_id": "bqr8y636",
215
+ "mode": null
216
+ }
217
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/training_state/optimizer_param_groups.json ADDED
@@ -0,0 +1,331 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "lr": 6.878069186763465e-05,
4
+ "betas": [
5
+ 0.95,
6
+ 0.999
7
+ ],
8
+ "eps": 1e-08,
9
+ "weight_decay": 1e-06,
10
+ "amsgrad": false,
11
+ "maximize": false,
12
+ "foreach": null,
13
+ "capturable": false,
14
+ "differentiable": false,
15
+ "fused": null,
16
+ "initial_lr": 0.0001,
17
+ "params": [
18
+ 0,
19
+ 1,
20
+ 2,
21
+ 3,
22
+ 4,
23
+ 5,
24
+ 6,
25
+ 7,
26
+ 8,
27
+ 9,
28
+ 10,
29
+ 11,
30
+ 12,
31
+ 13,
32
+ 14,
33
+ 15,
34
+ 16,
35
+ 17,
36
+ 18,
37
+ 19,
38
+ 20,
39
+ 21,
40
+ 22,
41
+ 23,
42
+ 24,
43
+ 25,
44
+ 26,
45
+ 27,
46
+ 28,
47
+ 29,
48
+ 30,
49
+ 31,
50
+ 32,
51
+ 33,
52
+ 34,
53
+ 35,
54
+ 36,
55
+ 37,
56
+ 38,
57
+ 39,
58
+ 40,
59
+ 41,
60
+ 42,
61
+ 43,
62
+ 44,
63
+ 45,
64
+ 46,
65
+ 47,
66
+ 48,
67
+ 49,
68
+ 50,
69
+ 51,
70
+ 52,
71
+ 53,
72
+ 54,
73
+ 55,
74
+ 56,
75
+ 57,
76
+ 58,
77
+ 59,
78
+ 60,
79
+ 61,
80
+ 62,
81
+ 63,
82
+ 64,
83
+ 65,
84
+ 66,
85
+ 67,
86
+ 68,
87
+ 69,
88
+ 70,
89
+ 71,
90
+ 72,
91
+ 73,
92
+ 74,
93
+ 75,
94
+ 76,
95
+ 77,
96
+ 78,
97
+ 79,
98
+ 80,
99
+ 81,
100
+ 82,
101
+ 83,
102
+ 84,
103
+ 85,
104
+ 86,
105
+ 87,
106
+ 88,
107
+ 89,
108
+ 90,
109
+ 91,
110
+ 92,
111
+ 93,
112
+ 94,
113
+ 95,
114
+ 96,
115
+ 97,
116
+ 98,
117
+ 99,
118
+ 100,
119
+ 101,
120
+ 102,
121
+ 103,
122
+ 104,
123
+ 105,
124
+ 106,
125
+ 107,
126
+ 108,
127
+ 109,
128
+ 110,
129
+ 111,
130
+ 112,
131
+ 113,
132
+ 114,
133
+ 115,
134
+ 116,
135
+ 117,
136
+ 118,
137
+ 119,
138
+ 120,
139
+ 121,
140
+ 122,
141
+ 123,
142
+ 124,
143
+ 125,
144
+ 126,
145
+ 127,
146
+ 128,
147
+ 129,
148
+ 130,
149
+ 131,
150
+ 132,
151
+ 133,
152
+ 134,
153
+ 135,
154
+ 136,
155
+ 137,
156
+ 138,
157
+ 139,
158
+ 140,
159
+ 141,
160
+ 142,
161
+ 143,
162
+ 144,
163
+ 145,
164
+ 146,
165
+ 147,
166
+ 148,
167
+ 149,
168
+ 150,
169
+ 151,
170
+ 152,
171
+ 153,
172
+ 154,
173
+ 155,
174
+ 156,
175
+ 157,
176
+ 158,
177
+ 159,
178
+ 160,
179
+ 161,
180
+ 162,
181
+ 163,
182
+ 164,
183
+ 165,
184
+ 166,
185
+ 167,
186
+ 168,
187
+ 169,
188
+ 170,
189
+ 171,
190
+ 172,
191
+ 173,
192
+ 174,
193
+ 175,
194
+ 176,
195
+ 177,
196
+ 178,
197
+ 179,
198
+ 180,
199
+ 181,
200
+ 182,
201
+ 183,
202
+ 184,
203
+ 185,
204
+ 186,
205
+ 187,
206
+ 188,
207
+ 189,
208
+ 190,
209
+ 191,
210
+ 192,
211
+ 193,
212
+ 194,
213
+ 195,
214
+ 196,
215
+ 197,
216
+ 198,
217
+ 199,
218
+ 200,
219
+ 201,
220
+ 202,
221
+ 203,
222
+ 204,
223
+ 205,
224
+ 206,
225
+ 207,
226
+ 208,
227
+ 209,
228
+ 210,
229
+ 211,
230
+ 212,
231
+ 213,
232
+ 214,
233
+ 215,
234
+ 216,
235
+ 217,
236
+ 218,
237
+ 219,
238
+ 220,
239
+ 221,
240
+ 222,
241
+ 223,
242
+ 224,
243
+ 225,
244
+ 226,
245
+ 227,
246
+ 228,
247
+ 229,
248
+ 230,
249
+ 231,
250
+ 232,
251
+ 233,
252
+ 234,
253
+ 235,
254
+ 236,
255
+ 237,
256
+ 238,
257
+ 239,
258
+ 240,
259
+ 241,
260
+ 242,
261
+ 243,
262
+ 244,
263
+ 245,
264
+ 246,
265
+ 247,
266
+ 248,
267
+ 249,
268
+ 250,
269
+ 251,
270
+ 252,
271
+ 253,
272
+ 254,
273
+ 255,
274
+ 256,
275
+ 257,
276
+ 258,
277
+ 259,
278
+ 260,
279
+ 261,
280
+ 262,
281
+ 263,
282
+ 264,
283
+ 265,
284
+ 266,
285
+ 267,
286
+ 268,
287
+ 269,
288
+ 270,
289
+ 271,
290
+ 272,
291
+ 273,
292
+ 274,
293
+ 275,
294
+ 276,
295
+ 277,
296
+ 278,
297
+ 279,
298
+ 280,
299
+ 281,
300
+ 282,
301
+ 283,
302
+ 284,
303
+ 285,
304
+ 286,
305
+ 287,
306
+ 288,
307
+ 289,
308
+ 290,
309
+ 291,
310
+ 292,
311
+ 293,
312
+ 294,
313
+ 295,
314
+ 296,
315
+ 297,
316
+ 298,
317
+ 299,
318
+ 300,
319
+ 301,
320
+ 302,
321
+ 303,
322
+ 304,
323
+ 305,
324
+ 306,
325
+ 307,
326
+ 308,
327
+ 309,
328
+ 310
329
+ ]
330
+ }
331
+ ]
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/training_state/optimizer_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:942763896eb0910b7fa8a563120278b58c6a5f5450b624c1c56f64e285916518
3
+ size 738026076
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/training_state/rng_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37ccf8dc123ae0aa4f67953719f34767f74aa76c9b1823a172edf3cfd645c5df
3
+ size 15708
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/training_state/scheduler_state.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_lrs": [
3
+ 0.0001
4
+ ],
5
+ "last_epoch": 340000,
6
+ "verbose": false,
7
+ "_step_count": 340001,
8
+ "_get_lr_called_within_step": false,
9
+ "_last_lr": [
10
+ 6.878069186763465e-05
11
+ ],
12
+ "lr_lambdas": [
13
+ null
14
+ ]
15
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/340000/training_state/training_step.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "step": 340000
3
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/pretrained_model/config.json ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "diffusion",
3
+ "n_obs_steps": 2,
4
+ "normalization_mapping": {
5
+ "VISUAL": "MEAN_STD",
6
+ "STATE": "MIN_MAX",
7
+ "ACTION": "MIN_MAX"
8
+ },
9
+ "input_features": {
10
+ "observation.state": {
11
+ "type": "STATE",
12
+ "shape": [
13
+ 6
14
+ ]
15
+ },
16
+ "observation.images.FrontCam": {
17
+ "type": "VISUAL",
18
+ "shape": [
19
+ 3,
20
+ 480,
21
+ 640
22
+ ]
23
+ },
24
+ "observation.images.TopCam": {
25
+ "type": "VISUAL",
26
+ "shape": [
27
+ 3,
28
+ 480,
29
+ 640
30
+ ]
31
+ },
32
+ "observation.images.WristCam": {
33
+ "type": "VISUAL",
34
+ "shape": [
35
+ 3,
36
+ 480,
37
+ 640
38
+ ]
39
+ }
40
+ },
41
+ "output_features": {
42
+ "action": {
43
+ "type": "ACTION",
44
+ "shape": [
45
+ 6
46
+ ]
47
+ }
48
+ },
49
+ "device": "cuda",
50
+ "use_amp": false,
51
+ "horizon": 16,
52
+ "n_action_steps": 8,
53
+ "drop_n_last_frames": 7,
54
+ "vision_backbone": "resnet50",
55
+ "crop_shape": [
56
+ 224,
57
+ 224
58
+ ],
59
+ "crop_is_random": false,
60
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
61
+ "use_group_norm": false,
62
+ "spatial_softmax_num_keypoints": 32,
63
+ "use_separate_rgb_encoder_per_camera": false,
64
+ "down_dims": [
65
+ 256,
66
+ 512,
67
+ 1024
68
+ ],
69
+ "kernel_size": 5,
70
+ "n_groups": 8,
71
+ "diffusion_step_embed_dim": 128,
72
+ "use_film_scale_modulation": true,
73
+ "noise_scheduler_type": "DDIM",
74
+ "num_train_timesteps": 100,
75
+ "beta_schedule": "squaredcos_cap_v2",
76
+ "beta_start": 0.0001,
77
+ "beta_end": 0.02,
78
+ "prediction_type": "epsilon",
79
+ "clip_sample": true,
80
+ "clip_sample_range": 1.0,
81
+ "num_inference_steps": 10,
82
+ "do_mask_loss_for_padding": false,
83
+ "optimizer_lr": 0.0001,
84
+ "optimizer_betas": [
85
+ 0.95,
86
+ 0.999
87
+ ],
88
+ "optimizer_eps": 1e-08,
89
+ "optimizer_weight_decay": 1e-06,
90
+ "scheduler_name": "cosine",
91
+ "scheduler_warmup_steps": 500,
92
+ "pre_resize_shape": [
93
+ 224,
94
+ 224
95
+ ],
96
+ "freeze_vision_backbone": false,
97
+ "use_ac": false,
98
+ "ac_threshold": 0.99,
99
+ "use_pcg": false,
100
+ "pcg_weight": 0.1,
101
+ "use_bid": false,
102
+ "bid_batch_size": 15,
103
+ "bid_mode_size": 3,
104
+ "bid_temporal_coherence": 0.5,
105
+ "weak_model_path": null,
106
+ "weak_model_config": null
107
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/pretrained_model/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab30ae7eb30be988093f96fd66145e2769d64a57d312e43c8e00bfe0eaab671a
3
+ size 369241872
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/pretrained_model/train_config.json ADDED
@@ -0,0 +1,217 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "shylee/so100_pengripG",
4
+ "root": "/SSD/LSY/lerobot_dataset/pengripG",
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
+ }
62
+ },
63
+ "revision": null,
64
+ "use_imagenet_stats": true,
65
+ "video_backend": "torchcodec"
66
+ },
67
+ "env": null,
68
+ "policy": {
69
+ "type": "diffusion",
70
+ "n_obs_steps": 2,
71
+ "normalization_mapping": {
72
+ "VISUAL": "MEAN_STD",
73
+ "STATE": "MIN_MAX",
74
+ "ACTION": "MIN_MAX"
75
+ },
76
+ "input_features": {
77
+ "observation.state": {
78
+ "type": "STATE",
79
+ "shape": [
80
+ 6
81
+ ]
82
+ },
83
+ "observation.images.FrontCam": {
84
+ "type": "VISUAL",
85
+ "shape": [
86
+ 3,
87
+ 480,
88
+ 640
89
+ ]
90
+ },
91
+ "observation.images.TopCam": {
92
+ "type": "VISUAL",
93
+ "shape": [
94
+ 3,
95
+ 480,
96
+ 640
97
+ ]
98
+ },
99
+ "observation.images.WristCam": {
100
+ "type": "VISUAL",
101
+ "shape": [
102
+ 3,
103
+ 480,
104
+ 640
105
+ ]
106
+ }
107
+ },
108
+ "output_features": {
109
+ "action": {
110
+ "type": "ACTION",
111
+ "shape": [
112
+ 6
113
+ ]
114
+ }
115
+ },
116
+ "device": "cuda",
117
+ "use_amp": false,
118
+ "horizon": 16,
119
+ "n_action_steps": 8,
120
+ "drop_n_last_frames": 7,
121
+ "vision_backbone": "resnet50",
122
+ "crop_shape": [
123
+ 224,
124
+ 224
125
+ ],
126
+ "crop_is_random": false,
127
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
128
+ "use_group_norm": false,
129
+ "spatial_softmax_num_keypoints": 32,
130
+ "use_separate_rgb_encoder_per_camera": false,
131
+ "down_dims": [
132
+ 256,
133
+ 512,
134
+ 1024
135
+ ],
136
+ "kernel_size": 5,
137
+ "n_groups": 8,
138
+ "diffusion_step_embed_dim": 128,
139
+ "use_film_scale_modulation": true,
140
+ "noise_scheduler_type": "DDIM",
141
+ "num_train_timesteps": 100,
142
+ "beta_schedule": "squaredcos_cap_v2",
143
+ "beta_start": 0.0001,
144
+ "beta_end": 0.02,
145
+ "prediction_type": "epsilon",
146
+ "clip_sample": true,
147
+ "clip_sample_range": 1.0,
148
+ "num_inference_steps": 10,
149
+ "do_mask_loss_for_padding": false,
150
+ "optimizer_lr": 0.0001,
151
+ "optimizer_betas": [
152
+ 0.95,
153
+ 0.999
154
+ ],
155
+ "optimizer_eps": 1e-08,
156
+ "optimizer_weight_decay": 1e-06,
157
+ "scheduler_name": "cosine",
158
+ "scheduler_warmup_steps": 500,
159
+ "pre_resize_shape": [
160
+ 224,
161
+ 224
162
+ ],
163
+ "freeze_vision_backbone": false,
164
+ "use_ac": false,
165
+ "ac_threshold": 0.99,
166
+ "use_pcg": false,
167
+ "pcg_weight": 0.1,
168
+ "use_bid": false,
169
+ "bid_batch_size": 15,
170
+ "bid_mode_size": 3,
171
+ "bid_temporal_coherence": 0.5,
172
+ "weak_model_path": null,
173
+ "weak_model_config": null
174
+ },
175
+ "output_dir": "/SSD/LSY/lerobot_model/DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
176
+ "job_name": "DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
177
+ "resume": true,
178
+ "seed": 1000,
179
+ "num_workers": 4,
180
+ "batch_size": 16,
181
+ "steps": 900000,
182
+ "eval_freq": 20000,
183
+ "log_freq": 200,
184
+ "save_checkpoint": true,
185
+ "save_freq": 20000,
186
+ "use_policy_training_preset": true,
187
+ "optimizer": {
188
+ "type": "adam",
189
+ "lr": 0.0001,
190
+ "weight_decay": 1e-06,
191
+ "grad_clip_norm": 10.0,
192
+ "betas": [
193
+ 0.95,
194
+ 0.999
195
+ ],
196
+ "eps": 1e-08
197
+ },
198
+ "scheduler": {
199
+ "type": "diffuser",
200
+ "num_warmup_steps": 500,
201
+ "name": "cosine"
202
+ },
203
+ "eval": {
204
+ "n_episodes": 50,
205
+ "batch_size": 50,
206
+ "use_async_envs": false
207
+ },
208
+ "wandb": {
209
+ "enable": true,
210
+ "disable_artifact": false,
211
+ "project": "lerobot",
212
+ "entity": null,
213
+ "notes": null,
214
+ "run_id": "bqr8y636",
215
+ "mode": null
216
+ }
217
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/training_state/optimizer_param_groups.json ADDED
@@ -0,0 +1,331 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "lr": 6.550066611157049e-05,
4
+ "betas": [
5
+ 0.95,
6
+ 0.999
7
+ ],
8
+ "eps": 1e-08,
9
+ "weight_decay": 1e-06,
10
+ "amsgrad": false,
11
+ "maximize": false,
12
+ "foreach": null,
13
+ "capturable": false,
14
+ "differentiable": false,
15
+ "fused": null,
16
+ "initial_lr": 0.0001,
17
+ "params": [
18
+ 0,
19
+ 1,
20
+ 2,
21
+ 3,
22
+ 4,
23
+ 5,
24
+ 6,
25
+ 7,
26
+ 8,
27
+ 9,
28
+ 10,
29
+ 11,
30
+ 12,
31
+ 13,
32
+ 14,
33
+ 15,
34
+ 16,
35
+ 17,
36
+ 18,
37
+ 19,
38
+ 20,
39
+ 21,
40
+ 22,
41
+ 23,
42
+ 24,
43
+ 25,
44
+ 26,
45
+ 27,
46
+ 28,
47
+ 29,
48
+ 30,
49
+ 31,
50
+ 32,
51
+ 33,
52
+ 34,
53
+ 35,
54
+ 36,
55
+ 37,
56
+ 38,
57
+ 39,
58
+ 40,
59
+ 41,
60
+ 42,
61
+ 43,
62
+ 44,
63
+ 45,
64
+ 46,
65
+ 47,
66
+ 48,
67
+ 49,
68
+ 50,
69
+ 51,
70
+ 52,
71
+ 53,
72
+ 54,
73
+ 55,
74
+ 56,
75
+ 57,
76
+ 58,
77
+ 59,
78
+ 60,
79
+ 61,
80
+ 62,
81
+ 63,
82
+ 64,
83
+ 65,
84
+ 66,
85
+ 67,
86
+ 68,
87
+ 69,
88
+ 70,
89
+ 71,
90
+ 72,
91
+ 73,
92
+ 74,
93
+ 75,
94
+ 76,
95
+ 77,
96
+ 78,
97
+ 79,
98
+ 80,
99
+ 81,
100
+ 82,
101
+ 83,
102
+ 84,
103
+ 85,
104
+ 86,
105
+ 87,
106
+ 88,
107
+ 89,
108
+ 90,
109
+ 91,
110
+ 92,
111
+ 93,
112
+ 94,
113
+ 95,
114
+ 96,
115
+ 97,
116
+ 98,
117
+ 99,
118
+ 100,
119
+ 101,
120
+ 102,
121
+ 103,
122
+ 104,
123
+ 105,
124
+ 106,
125
+ 107,
126
+ 108,
127
+ 109,
128
+ 110,
129
+ 111,
130
+ 112,
131
+ 113,
132
+ 114,
133
+ 115,
134
+ 116,
135
+ 117,
136
+ 118,
137
+ 119,
138
+ 120,
139
+ 121,
140
+ 122,
141
+ 123,
142
+ 124,
143
+ 125,
144
+ 126,
145
+ 127,
146
+ 128,
147
+ 129,
148
+ 130,
149
+ 131,
150
+ 132,
151
+ 133,
152
+ 134,
153
+ 135,
154
+ 136,
155
+ 137,
156
+ 138,
157
+ 139,
158
+ 140,
159
+ 141,
160
+ 142,
161
+ 143,
162
+ 144,
163
+ 145,
164
+ 146,
165
+ 147,
166
+ 148,
167
+ 149,
168
+ 150,
169
+ 151,
170
+ 152,
171
+ 153,
172
+ 154,
173
+ 155,
174
+ 156,
175
+ 157,
176
+ 158,
177
+ 159,
178
+ 160,
179
+ 161,
180
+ 162,
181
+ 163,
182
+ 164,
183
+ 165,
184
+ 166,
185
+ 167,
186
+ 168,
187
+ 169,
188
+ 170,
189
+ 171,
190
+ 172,
191
+ 173,
192
+ 174,
193
+ 175,
194
+ 176,
195
+ 177,
196
+ 178,
197
+ 179,
198
+ 180,
199
+ 181,
200
+ 182,
201
+ 183,
202
+ 184,
203
+ 185,
204
+ 186,
205
+ 187,
206
+ 188,
207
+ 189,
208
+ 190,
209
+ 191,
210
+ 192,
211
+ 193,
212
+ 194,
213
+ 195,
214
+ 196,
215
+ 197,
216
+ 198,
217
+ 199,
218
+ 200,
219
+ 201,
220
+ 202,
221
+ 203,
222
+ 204,
223
+ 205,
224
+ 206,
225
+ 207,
226
+ 208,
227
+ 209,
228
+ 210,
229
+ 211,
230
+ 212,
231
+ 213,
232
+ 214,
233
+ 215,
234
+ 216,
235
+ 217,
236
+ 218,
237
+ 219,
238
+ 220,
239
+ 221,
240
+ 222,
241
+ 223,
242
+ 224,
243
+ 225,
244
+ 226,
245
+ 227,
246
+ 228,
247
+ 229,
248
+ 230,
249
+ 231,
250
+ 232,
251
+ 233,
252
+ 234,
253
+ 235,
254
+ 236,
255
+ 237,
256
+ 238,
257
+ 239,
258
+ 240,
259
+ 241,
260
+ 242,
261
+ 243,
262
+ 244,
263
+ 245,
264
+ 246,
265
+ 247,
266
+ 248,
267
+ 249,
268
+ 250,
269
+ 251,
270
+ 252,
271
+ 253,
272
+ 254,
273
+ 255,
274
+ 256,
275
+ 257,
276
+ 258,
277
+ 259,
278
+ 260,
279
+ 261,
280
+ 262,
281
+ 263,
282
+ 264,
283
+ 265,
284
+ 266,
285
+ 267,
286
+ 268,
287
+ 269,
288
+ 270,
289
+ 271,
290
+ 272,
291
+ 273,
292
+ 274,
293
+ 275,
294
+ 276,
295
+ 277,
296
+ 278,
297
+ 279,
298
+ 280,
299
+ 281,
300
+ 282,
301
+ 283,
302
+ 284,
303
+ 285,
304
+ 286,
305
+ 287,
306
+ 288,
307
+ 289,
308
+ 290,
309
+ 291,
310
+ 292,
311
+ 293,
312
+ 294,
313
+ 295,
314
+ 296,
315
+ 297,
316
+ 298,
317
+ 299,
318
+ 300,
319
+ 301,
320
+ 302,
321
+ 303,
322
+ 304,
323
+ 305,
324
+ 306,
325
+ 307,
326
+ 308,
327
+ 309,
328
+ 310
329
+ ]
330
+ }
331
+ ]
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/training_state/optimizer_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c0b78a6d53dea135dcf1cd0f2f2ca939380801759e9b00aa519fa24dd875f36
3
+ size 738026076
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/training_state/rng_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68b85d369be8362b5b5e925e2da0a3e71387344dfb3f26f1cda1a9aed81bf66b
3
+ size 15708
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/training_state/scheduler_state.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_lrs": [
3
+ 0.0001
4
+ ],
5
+ "last_epoch": 360000,
6
+ "verbose": false,
7
+ "_step_count": 360001,
8
+ "_get_lr_called_within_step": false,
9
+ "_last_lr": [
10
+ 6.550066611157049e-05
11
+ ],
12
+ "lr_lambdas": [
13
+ null
14
+ ]
15
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/360000/training_state/training_step.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "step": 360000
3
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/pretrained_model/config.json ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "diffusion",
3
+ "n_obs_steps": 2,
4
+ "normalization_mapping": {
5
+ "VISUAL": "MEAN_STD",
6
+ "STATE": "MIN_MAX",
7
+ "ACTION": "MIN_MAX"
8
+ },
9
+ "input_features": {
10
+ "observation.state": {
11
+ "type": "STATE",
12
+ "shape": [
13
+ 6
14
+ ]
15
+ },
16
+ "observation.images.FrontCam": {
17
+ "type": "VISUAL",
18
+ "shape": [
19
+ 3,
20
+ 480,
21
+ 640
22
+ ]
23
+ },
24
+ "observation.images.TopCam": {
25
+ "type": "VISUAL",
26
+ "shape": [
27
+ 3,
28
+ 480,
29
+ 640
30
+ ]
31
+ },
32
+ "observation.images.WristCam": {
33
+ "type": "VISUAL",
34
+ "shape": [
35
+ 3,
36
+ 480,
37
+ 640
38
+ ]
39
+ }
40
+ },
41
+ "output_features": {
42
+ "action": {
43
+ "type": "ACTION",
44
+ "shape": [
45
+ 6
46
+ ]
47
+ }
48
+ },
49
+ "device": "cuda",
50
+ "use_amp": false,
51
+ "horizon": 16,
52
+ "n_action_steps": 8,
53
+ "drop_n_last_frames": 7,
54
+ "vision_backbone": "resnet50",
55
+ "crop_shape": [
56
+ 224,
57
+ 224
58
+ ],
59
+ "crop_is_random": false,
60
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
61
+ "use_group_norm": false,
62
+ "spatial_softmax_num_keypoints": 32,
63
+ "use_separate_rgb_encoder_per_camera": false,
64
+ "down_dims": [
65
+ 256,
66
+ 512,
67
+ 1024
68
+ ],
69
+ "kernel_size": 5,
70
+ "n_groups": 8,
71
+ "diffusion_step_embed_dim": 128,
72
+ "use_film_scale_modulation": true,
73
+ "noise_scheduler_type": "DDIM",
74
+ "num_train_timesteps": 100,
75
+ "beta_schedule": "squaredcos_cap_v2",
76
+ "beta_start": 0.0001,
77
+ "beta_end": 0.02,
78
+ "prediction_type": "epsilon",
79
+ "clip_sample": true,
80
+ "clip_sample_range": 1.0,
81
+ "num_inference_steps": 10,
82
+ "do_mask_loss_for_padding": false,
83
+ "optimizer_lr": 0.0001,
84
+ "optimizer_betas": [
85
+ 0.95,
86
+ 0.999
87
+ ],
88
+ "optimizer_eps": 1e-08,
89
+ "optimizer_weight_decay": 1e-06,
90
+ "scheduler_name": "cosine",
91
+ "scheduler_warmup_steps": 500,
92
+ "pre_resize_shape": [
93
+ 224,
94
+ 224
95
+ ],
96
+ "freeze_vision_backbone": false,
97
+ "use_ac": false,
98
+ "ac_threshold": 0.99,
99
+ "use_pcg": false,
100
+ "pcg_weight": 0.1,
101
+ "use_bid": false,
102
+ "bid_batch_size": 15,
103
+ "bid_mode_size": 3,
104
+ "bid_temporal_coherence": 0.5,
105
+ "weak_model_path": null,
106
+ "weak_model_config": null
107
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/pretrained_model/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7866c809900a0b29765a29297bf5a2968f1570997e3b62a8607848a6696859e
3
+ size 369241872
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/pretrained_model/train_config.json ADDED
@@ -0,0 +1,217 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "shylee/so100_pengripG",
4
+ "root": "/SSD/LSY/lerobot_dataset/pengripG",
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
+ }
62
+ },
63
+ "revision": null,
64
+ "use_imagenet_stats": true,
65
+ "video_backend": "torchcodec"
66
+ },
67
+ "env": null,
68
+ "policy": {
69
+ "type": "diffusion",
70
+ "n_obs_steps": 2,
71
+ "normalization_mapping": {
72
+ "VISUAL": "MEAN_STD",
73
+ "STATE": "MIN_MAX",
74
+ "ACTION": "MIN_MAX"
75
+ },
76
+ "input_features": {
77
+ "observation.state": {
78
+ "type": "STATE",
79
+ "shape": [
80
+ 6
81
+ ]
82
+ },
83
+ "observation.images.FrontCam": {
84
+ "type": "VISUAL",
85
+ "shape": [
86
+ 3,
87
+ 480,
88
+ 640
89
+ ]
90
+ },
91
+ "observation.images.TopCam": {
92
+ "type": "VISUAL",
93
+ "shape": [
94
+ 3,
95
+ 480,
96
+ 640
97
+ ]
98
+ },
99
+ "observation.images.WristCam": {
100
+ "type": "VISUAL",
101
+ "shape": [
102
+ 3,
103
+ 480,
104
+ 640
105
+ ]
106
+ }
107
+ },
108
+ "output_features": {
109
+ "action": {
110
+ "type": "ACTION",
111
+ "shape": [
112
+ 6
113
+ ]
114
+ }
115
+ },
116
+ "device": "cuda",
117
+ "use_amp": false,
118
+ "horizon": 16,
119
+ "n_action_steps": 8,
120
+ "drop_n_last_frames": 7,
121
+ "vision_backbone": "resnet50",
122
+ "crop_shape": [
123
+ 224,
124
+ 224
125
+ ],
126
+ "crop_is_random": false,
127
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
128
+ "use_group_norm": false,
129
+ "spatial_softmax_num_keypoints": 32,
130
+ "use_separate_rgb_encoder_per_camera": false,
131
+ "down_dims": [
132
+ 256,
133
+ 512,
134
+ 1024
135
+ ],
136
+ "kernel_size": 5,
137
+ "n_groups": 8,
138
+ "diffusion_step_embed_dim": 128,
139
+ "use_film_scale_modulation": true,
140
+ "noise_scheduler_type": "DDIM",
141
+ "num_train_timesteps": 100,
142
+ "beta_schedule": "squaredcos_cap_v2",
143
+ "beta_start": 0.0001,
144
+ "beta_end": 0.02,
145
+ "prediction_type": "epsilon",
146
+ "clip_sample": true,
147
+ "clip_sample_range": 1.0,
148
+ "num_inference_steps": 10,
149
+ "do_mask_loss_for_padding": false,
150
+ "optimizer_lr": 0.0001,
151
+ "optimizer_betas": [
152
+ 0.95,
153
+ 0.999
154
+ ],
155
+ "optimizer_eps": 1e-08,
156
+ "optimizer_weight_decay": 1e-06,
157
+ "scheduler_name": "cosine",
158
+ "scheduler_warmup_steps": 500,
159
+ "pre_resize_shape": [
160
+ 224,
161
+ 224
162
+ ],
163
+ "freeze_vision_backbone": false,
164
+ "use_ac": false,
165
+ "ac_threshold": 0.99,
166
+ "use_pcg": false,
167
+ "pcg_weight": 0.1,
168
+ "use_bid": false,
169
+ "bid_batch_size": 15,
170
+ "bid_mode_size": 3,
171
+ "bid_temporal_coherence": 0.5,
172
+ "weak_model_path": null,
173
+ "weak_model_config": null
174
+ },
175
+ "output_dir": "/SSD/LSY/lerobot_model/DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
176
+ "job_name": "DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain",
177
+ "resume": true,
178
+ "seed": 1000,
179
+ "num_workers": 4,
180
+ "batch_size": 16,
181
+ "steps": 900000,
182
+ "eval_freq": 20000,
183
+ "log_freq": 200,
184
+ "save_checkpoint": true,
185
+ "save_freq": 20000,
186
+ "use_policy_training_preset": true,
187
+ "optimizer": {
188
+ "type": "adam",
189
+ "lr": 0.0001,
190
+ "weight_decay": 1e-06,
191
+ "grad_clip_norm": 10.0,
192
+ "betas": [
193
+ 0.95,
194
+ 0.999
195
+ ],
196
+ "eps": 1e-08
197
+ },
198
+ "scheduler": {
199
+ "type": "diffuser",
200
+ "num_warmup_steps": 500,
201
+ "name": "cosine"
202
+ },
203
+ "eval": {
204
+ "n_episodes": 50,
205
+ "batch_size": 50,
206
+ "use_async_envs": false
207
+ },
208
+ "wandb": {
209
+ "enable": true,
210
+ "disable_artifact": false,
211
+ "project": "lerobot",
212
+ "entity": null,
213
+ "notes": null,
214
+ "run_id": "bqr8y636",
215
+ "mode": null
216
+ }
217
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/training_state/optimizer_param_groups.json ADDED
@@ -0,0 +1,331 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "lr": 6.214503872396447e-05,
4
+ "betas": [
5
+ 0.95,
6
+ 0.999
7
+ ],
8
+ "eps": 1e-08,
9
+ "weight_decay": 1e-06,
10
+ "amsgrad": false,
11
+ "maximize": false,
12
+ "foreach": null,
13
+ "capturable": false,
14
+ "differentiable": false,
15
+ "fused": null,
16
+ "initial_lr": 0.0001,
17
+ "params": [
18
+ 0,
19
+ 1,
20
+ 2,
21
+ 3,
22
+ 4,
23
+ 5,
24
+ 6,
25
+ 7,
26
+ 8,
27
+ 9,
28
+ 10,
29
+ 11,
30
+ 12,
31
+ 13,
32
+ 14,
33
+ 15,
34
+ 16,
35
+ 17,
36
+ 18,
37
+ 19,
38
+ 20,
39
+ 21,
40
+ 22,
41
+ 23,
42
+ 24,
43
+ 25,
44
+ 26,
45
+ 27,
46
+ 28,
47
+ 29,
48
+ 30,
49
+ 31,
50
+ 32,
51
+ 33,
52
+ 34,
53
+ 35,
54
+ 36,
55
+ 37,
56
+ 38,
57
+ 39,
58
+ 40,
59
+ 41,
60
+ 42,
61
+ 43,
62
+ 44,
63
+ 45,
64
+ 46,
65
+ 47,
66
+ 48,
67
+ 49,
68
+ 50,
69
+ 51,
70
+ 52,
71
+ 53,
72
+ 54,
73
+ 55,
74
+ 56,
75
+ 57,
76
+ 58,
77
+ 59,
78
+ 60,
79
+ 61,
80
+ 62,
81
+ 63,
82
+ 64,
83
+ 65,
84
+ 66,
85
+ 67,
86
+ 68,
87
+ 69,
88
+ 70,
89
+ 71,
90
+ 72,
91
+ 73,
92
+ 74,
93
+ 75,
94
+ 76,
95
+ 77,
96
+ 78,
97
+ 79,
98
+ 80,
99
+ 81,
100
+ 82,
101
+ 83,
102
+ 84,
103
+ 85,
104
+ 86,
105
+ 87,
106
+ 88,
107
+ 89,
108
+ 90,
109
+ 91,
110
+ 92,
111
+ 93,
112
+ 94,
113
+ 95,
114
+ 96,
115
+ 97,
116
+ 98,
117
+ 99,
118
+ 100,
119
+ 101,
120
+ 102,
121
+ 103,
122
+ 104,
123
+ 105,
124
+ 106,
125
+ 107,
126
+ 108,
127
+ 109,
128
+ 110,
129
+ 111,
130
+ 112,
131
+ 113,
132
+ 114,
133
+ 115,
134
+ 116,
135
+ 117,
136
+ 118,
137
+ 119,
138
+ 120,
139
+ 121,
140
+ 122,
141
+ 123,
142
+ 124,
143
+ 125,
144
+ 126,
145
+ 127,
146
+ 128,
147
+ 129,
148
+ 130,
149
+ 131,
150
+ 132,
151
+ 133,
152
+ 134,
153
+ 135,
154
+ 136,
155
+ 137,
156
+ 138,
157
+ 139,
158
+ 140,
159
+ 141,
160
+ 142,
161
+ 143,
162
+ 144,
163
+ 145,
164
+ 146,
165
+ 147,
166
+ 148,
167
+ 149,
168
+ 150,
169
+ 151,
170
+ 152,
171
+ 153,
172
+ 154,
173
+ 155,
174
+ 156,
175
+ 157,
176
+ 158,
177
+ 159,
178
+ 160,
179
+ 161,
180
+ 162,
181
+ 163,
182
+ 164,
183
+ 165,
184
+ 166,
185
+ 167,
186
+ 168,
187
+ 169,
188
+ 170,
189
+ 171,
190
+ 172,
191
+ 173,
192
+ 174,
193
+ 175,
194
+ 176,
195
+ 177,
196
+ 178,
197
+ 179,
198
+ 180,
199
+ 181,
200
+ 182,
201
+ 183,
202
+ 184,
203
+ 185,
204
+ 186,
205
+ 187,
206
+ 188,
207
+ 189,
208
+ 190,
209
+ 191,
210
+ 192,
211
+ 193,
212
+ 194,
213
+ 195,
214
+ 196,
215
+ 197,
216
+ 198,
217
+ 199,
218
+ 200,
219
+ 201,
220
+ 202,
221
+ 203,
222
+ 204,
223
+ 205,
224
+ 206,
225
+ 207,
226
+ 208,
227
+ 209,
228
+ 210,
229
+ 211,
230
+ 212,
231
+ 213,
232
+ 214,
233
+ 215,
234
+ 216,
235
+ 217,
236
+ 218,
237
+ 219,
238
+ 220,
239
+ 221,
240
+ 222,
241
+ 223,
242
+ 224,
243
+ 225,
244
+ 226,
245
+ 227,
246
+ 228,
247
+ 229,
248
+ 230,
249
+ 231,
250
+ 232,
251
+ 233,
252
+ 234,
253
+ 235,
254
+ 236,
255
+ 237,
256
+ 238,
257
+ 239,
258
+ 240,
259
+ 241,
260
+ 242,
261
+ 243,
262
+ 244,
263
+ 245,
264
+ 246,
265
+ 247,
266
+ 248,
267
+ 249,
268
+ 250,
269
+ 251,
270
+ 252,
271
+ 253,
272
+ 254,
273
+ 255,
274
+ 256,
275
+ 257,
276
+ 258,
277
+ 259,
278
+ 260,
279
+ 261,
280
+ 262,
281
+ 263,
282
+ 264,
283
+ 265,
284
+ 266,
285
+ 267,
286
+ 268,
287
+ 269,
288
+ 270,
289
+ 271,
290
+ 272,
291
+ 273,
292
+ 274,
293
+ 275,
294
+ 276,
295
+ 277,
296
+ 278,
297
+ 279,
298
+ 280,
299
+ 281,
300
+ 282,
301
+ 283,
302
+ 284,
303
+ 285,
304
+ 286,
305
+ 287,
306
+ 288,
307
+ 289,
308
+ 290,
309
+ 291,
310
+ 292,
311
+ 293,
312
+ 294,
313
+ 295,
314
+ 296,
315
+ 297,
316
+ 298,
317
+ 299,
318
+ 300,
319
+ 301,
320
+ 302,
321
+ 303,
322
+ 304,
323
+ 305,
324
+ 306,
325
+ 307,
326
+ 308,
327
+ 309,
328
+ 310
329
+ ]
330
+ }
331
+ ]
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/training_state/optimizer_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc6f7e3930b1eefa2f2b586fa6a4136ea76b18ece12b67bfe5024d6e44835f50
3
+ size 738026076
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/training_state/rng_state.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a0de54a7a1612a7bf707e4e2733964d583a5fdcec574a2d38c80ff8ac41bfa1
3
+ size 15708
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/training_state/scheduler_state.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_lrs": [
3
+ 0.0001
4
+ ],
5
+ "last_epoch": 380000,
6
+ "verbose": false,
7
+ "_step_count": 380001,
8
+ "_get_lr_called_within_step": false,
9
+ "_last_lr": [
10
+ 6.214503872396447e-05
11
+ ],
12
+ "lr_lambdas": [
13
+ null
14
+ ]
15
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/380000/training_state/training_step.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "step": 380000
3
+ }
DP_pengripG_input224_freeze0_horizon16_transform0_1e-4_longtrain/checkpoints/400000/pretrained_model/config.json ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "diffusion",
3
+ "n_obs_steps": 2,
4
+ "normalization_mapping": {
5
+ "VISUAL": "MEAN_STD",
6
+ "STATE": "MIN_MAX",
7
+ "ACTION": "MIN_MAX"
8
+ },
9
+ "input_features": {
10
+ "observation.state": {
11
+ "type": "STATE",
12
+ "shape": [
13
+ 6
14
+ ]
15
+ },
16
+ "observation.images.FrontCam": {
17
+ "type": "VISUAL",
18
+ "shape": [
19
+ 3,
20
+ 480,
21
+ 640
22
+ ]
23
+ },
24
+ "observation.images.TopCam": {
25
+ "type": "VISUAL",
26
+ "shape": [
27
+ 3,
28
+ 480,
29
+ 640
30
+ ]
31
+ },
32
+ "observation.images.WristCam": {
33
+ "type": "VISUAL",
34
+ "shape": [
35
+ 3,
36
+ 480,
37
+ 640
38
+ ]
39
+ }
40
+ },
41
+ "output_features": {
42
+ "action": {
43
+ "type": "ACTION",
44
+ "shape": [
45
+ 6
46
+ ]
47
+ }
48
+ },
49
+ "device": "cuda",
50
+ "use_amp": false,
51
+ "horizon": 16,
52
+ "n_action_steps": 8,
53
+ "drop_n_last_frames": 7,
54
+ "vision_backbone": "resnet50",
55
+ "crop_shape": [
56
+ 224,
57
+ 224
58
+ ],
59
+ "crop_is_random": false,
60
+ "pretrained_backbone_weights": "ResNet50_Weights.IMAGENET1K_V1",
61
+ "use_group_norm": false,
62
+ "spatial_softmax_num_keypoints": 32,
63
+ "use_separate_rgb_encoder_per_camera": false,
64
+ "down_dims": [
65
+ 256,
66
+ 512,
67
+ 1024
68
+ ],
69
+ "kernel_size": 5,
70
+ "n_groups": 8,
71
+ "diffusion_step_embed_dim": 128,
72
+ "use_film_scale_modulation": true,
73
+ "noise_scheduler_type": "DDIM",
74
+ "num_train_timesteps": 100,
75
+ "beta_schedule": "squaredcos_cap_v2",
76
+ "beta_start": 0.0001,
77
+ "beta_end": 0.02,
78
+ "prediction_type": "epsilon",
79
+ "clip_sample": true,
80
+ "clip_sample_range": 1.0,
81
+ "num_inference_steps": 10,
82
+ "do_mask_loss_for_padding": false,
83
+ "optimizer_lr": 0.0001,
84
+ "optimizer_betas": [
85
+ 0.95,
86
+ 0.999
87
+ ],
88
+ "optimizer_eps": 1e-08,
89
+ "optimizer_weight_decay": 1e-06,
90
+ "scheduler_name": "cosine",
91
+ "scheduler_warmup_steps": 500,
92
+ "pre_resize_shape": [
93
+ 224,
94
+ 224
95
+ ],
96
+ "freeze_vision_backbone": false,
97
+ "use_ac": false,
98
+ "ac_threshold": 0.99,
99
+ "use_pcg": false,
100
+ "pcg_weight": 0.1,
101
+ "use_bid": false,
102
+ "bid_batch_size": 15,
103
+ "bid_mode_size": 3,
104
+ "bid_temporal_coherence": 0.5,
105
+ "weak_model_path": null,
106
+ "weak_model_config": null
107
+ }