Hugging Face's logo

gaspardthrl
/
walleed-diffusion-A3

Safetensors
Model card Files
xet
 Community
gaspardthrl commited on
Commit
73e90c9
·
verified ·
1 Parent(s): 6099ce4

Upload folder using huggingface_hub

Browse files
Files changed (7) hide show
  1. 040000/pretrained_model/config.json +108 -0
  2. 040000/pretrained_model/model.safetensors +3 -0
  3. 040000/pretrained_model/policy_postprocessor.json +38 -0
  4. 040000/pretrained_model/policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
  5. 040000/pretrained_model/policy_preprocessor.json +73 -0
  6. 040000/pretrained_model/policy_preprocessor_step_4_normalizer_processor.safetensors +3 -0
  7. 040000/pretrained_model/train_config.json +225 -0
040000/pretrained_model/config.json ADDED
@@ -0,0 +1,108 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "diffusion",
3
+ "n_obs_steps": 1,
4
+ "input_features": {
5
+ "observation.state": {
6
+ "type": "STATE",
7
+ "shape": [
8
+ 6
9
+ ]
10
+ },
11
+ "observation.images.front": {
12
+ "type": "VISUAL",
13
+ "shape": [
14
+ 3,
15
+ 480,
16
+ 640
17
+ ]
18
+ }
19
+ },
20
+ "output_features": {
21
+ "action": {
22
+ "type": "ACTION",
23
+ "shape": [
24
+ 6
25
+ ]
26
+ }
27
+ },
28
+ "device": "cuda",
29
+ "use_amp": false,
30
+ "use_peft": false,
31
+ "push_to_hub": true,
32
+ "repo_id": "gaspardthrl/diffusion_A3",
33
+ "private": null,
34
+ "tags": null,
35
+ "license": null,
36
+ "pretrained_path": null,
37
+ "horizon": 32,
38
+ "n_action_steps": 16,
39
+ "normalization_mapping": {
40
+ "VISUAL": "MEAN_STD",
41
+ "STATE": "MIN_MAX",
42
+ "ACTION": "MEAN_STD"
43
+ },
44
+ "drop_n_last_frames": 16,
45
+ "use_relative_actions": true,
46
+ "relative_exclude_joints": [
47
+ "gripper"
48
+ ],
49
+ "action_feature_names": [
50
+ "shoulder_pan.pos",
51
+ "shoulder_lift.pos",
52
+ "elbow_flex.pos",
53
+ "wrist_flex.pos",
54
+ "wrist_roll.pos",
55
+ "gripper.pos"
56
+ ],
57
+ "proprio_dropout": 0.3,
58
+ "state_indices": null,
59
+ "action_normalization_mode": "MEAN_STD",
60
+ "use_eef_actions": false,
61
+ "eef_poses_path": null,
62
+ "eef_stats_path": null,
63
+ "eef_action_dim": 10,
64
+ "vision_backbone": "resnet18",
65
+ "resize_shape": [
66
+ 96,
67
+ 96
68
+ ],
69
+ "crop_ratio": 1.0,
70
+ "crop_shape": null,
71
+ "crop_is_random": true,
72
+ "pretrained_backbone_weights": null,
73
+ "use_group_norm": true,
74
+ "spatial_softmax_num_keypoints": 32,
75
+ "use_separate_rgb_encoder_per_camera": false,
76
+ "image_grayworld": true,
77
+ "image_grayscale": false,
78
+ "down_dims": [
79
+ 512,
80
+ 1024,
81
+ 2048
82
+ ],
83
+ "kernel_size": 5,
84
+ "n_groups": 8,
85
+ "diffusion_step_embed_dim": 128,
86
+ "use_film_scale_modulation": true,
87
+ "noise_scheduler_type": "DDPM",
88
+ "num_train_timesteps": 100,
89
+ "beta_schedule": "squaredcos_cap_v2",
90
+ "beta_start": 0.0001,
91
+ "beta_end": 0.02,
92
+ "prediction_type": "epsilon",
93
+ "clip_sample": true,
94
+ "clip_sample_range": 3.0,
95
+ "num_inference_steps": null,
96
+ "compile_model": false,
97
+ "compile_mode": "reduce-overhead",
98
+ "do_mask_loss_for_padding": false,
99
+ "optimizer_lr": 0.0001,
100
+ "optimizer_betas": [
101
+ 0.95,
102
+ 0.999
103
+ ],
104
+ "optimizer_eps": 1e-08,
105
+ "optimizer_weight_decay": 1e-06,
106
+ "scheduler_name": "cosine",
107
+ "scheduler_warmup_steps": 500
108
+ }
040000/pretrained_model/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b07043285141a59c25bce41e94afff08c9423368658d98293b13fb2cad2beb39
3
+ size 1043808136
040000/pretrained_model/policy_postprocessor.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "policy_postprocessor",
3
+ "steps": [
4
+ {
5
+ "registry_name": "unnormalizer_processor",
6
+ "config": {
7
+ "eps": 1e-08,
8
+ "features": {
9
+ "action": {
10
+ "type": "ACTION",
11
+ "shape": [
12
+ 6
13
+ ]
14
+ }
15
+ },
16
+ "norm_map": {
17
+ "VISUAL": "MEAN_STD",
18
+ "STATE": "MIN_MAX",
19
+ "ACTION": "MEAN_STD"
20
+ }
21
+ },
22
+ "state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
23
+ },
24
+ {
25
+ "registry_name": "absolute_actions_processor",
26
+ "config": {
27
+ "enabled": true
28
+ }
29
+ },
30
+ {
31
+ "registry_name": "device_processor",
32
+ "config": {
33
+ "device": "cpu",
34
+ "float_dtype": null
35
+ }
36
+ }
37
+ ]
38
+ }
040000/pretrained_model/policy_postprocessor_step_0_unnormalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:56c06ae2b2aae3ef1cb97def3966892ec0cedc551dd3028981658eee9408aad8
3
+ size 6580
040000/pretrained_model/policy_preprocessor.json ADDED
@@ -0,0 +1,73 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "policy_preprocessor",
3
+ "steps": [
4
+ {
5
+ "registry_name": "rename_observations_processor",
6
+ "config": {
7
+ "rename_map": {}
8
+ }
9
+ },
10
+ {
11
+ "registry_name": "to_batch_processor",
12
+ "config": {}
13
+ },
14
+ {
15
+ "registry_name": "device_processor",
16
+ "config": {
17
+ "device": "cuda",
18
+ "float_dtype": null
19
+ }
20
+ },
21
+ {
22
+ "registry_name": "delta_actions_processor",
23
+ "config": {
24
+ "enabled": true,
25
+ "exclude_joints": [
26
+ "gripper"
27
+ ],
28
+ "action_names": [
29
+ "shoulder_pan.pos",
30
+ "shoulder_lift.pos",
31
+ "elbow_flex.pos",
32
+ "wrist_flex.pos",
33
+ "wrist_roll.pos",
34
+ "gripper.pos"
35
+ ]
36
+ }
37
+ },
38
+ {
39
+ "registry_name": "normalizer_processor",
40
+ "config": {
41
+ "eps": 1e-08,
42
+ "features": {
43
+ "observation.state": {
44
+ "type": "STATE",
45
+ "shape": [
46
+ 6
47
+ ]
48
+ },
49
+ "observation.images.front": {
50
+ "type": "VISUAL",
51
+ "shape": [
52
+ 3,
53
+ 480,
54
+ 640
55
+ ]
56
+ },
57
+ "action": {
58
+ "type": "ACTION",
59
+ "shape": [
60
+ 6
61
+ ]
62
+ }
63
+ },
64
+ "norm_map": {
65
+ "VISUAL": "MEAN_STD",
66
+ "STATE": "MIN_MAX",
67
+ "ACTION": "MEAN_STD"
68
+ }
69
+ },
70
+ "state_file": "policy_preprocessor_step_4_normalizer_processor.safetensors"
71
+ }
72
+ ]
73
+ }
040000/pretrained_model/policy_preprocessor_step_4_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:56c06ae2b2aae3ef1cb97def3966892ec0cedc551dd3028981658eee9408aad8
3
+ size 6580
040000/pretrained_model/train_config.json ADDED
@@ -0,0 +1,225 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "gaspardthrl/walleed_fold_combined",
4
+ "root": null,
5
+ "episodes": null,
6
+ "image_transforms": {
7
+ "enable": true,
8
+ "max_num_transforms": 4,
9
+ "random_order": true,
10
+ "tfs": {
11
+ "brightness": {
12
+ "weight": 1.0,
13
+ "type": "ColorJitter",
14
+ "kwargs": {
15
+ "brightness": [
16
+ 0.7,
17
+ 1.3
18
+ ]
19
+ }
20
+ },
21
+ "contrast": {
22
+ "weight": 1.0,
23
+ "type": "ColorJitter",
24
+ "kwargs": {
25
+ "contrast": [
26
+ 0.7,
27
+ 1.3
28
+ ]
29
+ }
30
+ },
31
+ "warmth": {
32
+ "weight": 1.0,
33
+ "type": "WarmthJitter",
34
+ "kwargs": {
35
+ "warmth": [
36
+ -0.15,
37
+ 0.15
38
+ ]
39
+ }
40
+ },
41
+ "affine": {
42
+ "weight": 1.0,
43
+ "type": "RandomAffine",
44
+ "kwargs": {
45
+ "degrees": [
46
+ -8,
47
+ 8
48
+ ],
49
+ "translate": [
50
+ 0.12,
51
+ 0.12
52
+ ]
53
+ }
54
+ }
55
+ }
56
+ },
57
+ "revision": null,
58
+ "use_imagenet_stats": true,
59
+ "stats_override_path": "./data/meta/relative_stats.json",
60
+ "video_backend": "pyav",
61
+ "return_uint8": false,
62
+ "streaming": false
63
+ },
64
+ "env": null,
65
+ "policy": {
66
+ "type": "diffusion",
67
+ "n_obs_steps": 1,
68
+ "input_features": {
69
+ "observation.state": {
70
+ "type": "STATE",
71
+ "shape": [
72
+ 6
73
+ ]
74
+ },
75
+ "observation.images.front": {
76
+ "type": "VISUAL",
77
+ "shape": [
78
+ 3,
79
+ 480,
80
+ 640
81
+ ]
82
+ }
83
+ },
84
+ "output_features": {
85
+ "action": {
86
+ "type": "ACTION",
87
+ "shape": [
88
+ 6
89
+ ]
90
+ }
91
+ },
92
+ "device": "cuda",
93
+ "use_amp": false,
94
+ "use_peft": false,
95
+ "push_to_hub": true,
96
+ "repo_id": "gaspardthrl/diffusion_A3",
97
+ "private": null,
98
+ "tags": null,
99
+ "license": null,
100
+ "pretrained_path": null,
101
+ "horizon": 32,
102
+ "n_action_steps": 16,
103
+ "normalization_mapping": {
104
+ "VISUAL": "MEAN_STD",
105
+ "STATE": "MIN_MAX",
106
+ "ACTION": "MEAN_STD"
107
+ },
108
+ "drop_n_last_frames": 16,
109
+ "use_relative_actions": true,
110
+ "relative_exclude_joints": [
111
+ "gripper"
112
+ ],
113
+ "action_feature_names": [
114
+ "shoulder_pan.pos",
115
+ "shoulder_lift.pos",
116
+ "elbow_flex.pos",
117
+ "wrist_flex.pos",
118
+ "wrist_roll.pos",
119
+ "gripper.pos"
120
+ ],
121
+ "proprio_dropout": 0.3,
122
+ "state_indices": null,
123
+ "action_normalization_mode": "MEAN_STD",
124
+ "use_eef_actions": false,
125
+ "eef_poses_path": null,
126
+ "eef_stats_path": null,
127
+ "eef_action_dim": 10,
128
+ "vision_backbone": "resnet18",
129
+ "resize_shape": [
130
+ 96,
131
+ 96
132
+ ],
133
+ "crop_ratio": 1.0,
134
+ "crop_shape": null,
135
+ "crop_is_random": true,
136
+ "pretrained_backbone_weights": null,
137
+ "use_group_norm": true,
138
+ "spatial_softmax_num_keypoints": 32,
139
+ "use_separate_rgb_encoder_per_camera": false,
140
+ "image_grayworld": true,
141
+ "image_grayscale": false,
142
+ "down_dims": [
143
+ 512,
144
+ 1024,
145
+ 2048
146
+ ],
147
+ "kernel_size": 5,
148
+ "n_groups": 8,
149
+ "diffusion_step_embed_dim": 128,
150
+ "use_film_scale_modulation": true,
151
+ "noise_scheduler_type": "DDPM",
152
+ "num_train_timesteps": 100,
153
+ "beta_schedule": "squaredcos_cap_v2",
154
+ "beta_start": 0.0001,
155
+ "beta_end": 0.02,
156
+ "prediction_type": "epsilon",
157
+ "clip_sample": true,
158
+ "clip_sample_range": 3.0,
159
+ "num_inference_steps": null,
160
+ "compile_model": false,
161
+ "compile_mode": "reduce-overhead",
162
+ "do_mask_loss_for_padding": false,
163
+ "optimizer_lr": 0.0001,
164
+ "optimizer_betas": [
165
+ 0.95,
166
+ 0.999
167
+ ],
168
+ "optimizer_eps": 1e-08,
169
+ "optimizer_weight_decay": 1e-06,
170
+ "scheduler_name": "cosine",
171
+ "scheduler_warmup_steps": 500
172
+ },
173
+ "reward_model": null,
174
+ "output_dir": "outputs/diffusion_A3_20260517_133735",
175
+ "job_name": "diffusion-A3-proprio-dropout",
176
+ "resume": false,
177
+ "seed": 1000,
178
+ "cudnn_deterministic": false,
179
+ "num_workers": 8,
180
+ "batch_size": 128,
181
+ "prefetch_factor": 4,
182
+ "persistent_workers": true,
183
+ "steps": 100000,
184
+ "eval_freq": 20000,
185
+ "log_freq": 200,
186
+ "tolerance_s": 0.0001,
187
+ "save_checkpoint": true,
188
+ "save_freq": 20000,
189
+ "use_policy_training_preset": true,
190
+ "optimizer": {
191
+ "type": "adam",
192
+ "lr": 0.0001,
193
+ "weight_decay": 1e-06,
194
+ "grad_clip_norm": 10.0,
195
+ "betas": [
196
+ 0.95,
197
+ 0.999
198
+ ],
199
+ "eps": 1e-08
200
+ },
201
+ "scheduler": {
202
+ "type": "diffuser",
203
+ "num_warmup_steps": 500,
204
+ "name": "cosine"
205
+ },
206
+ "eval": {
207
+ "n_episodes": 50,
208
+ "batch_size": 5,
209
+ "use_async_envs": true
210
+ },
211
+ "wandb": {
212
+ "enable": true,
213
+ "disable_artifact": false,
214
+ "project": "diffusing",
215
+ "entity": null,
216
+ "notes": "A3 proprio dropout p=0.3, relative actions, MEAN_STD action norm, ResNet-18, 100000 steps",
217
+ "run_id": "7ycy74ml",
218
+ "mode": null,
219
+ "add_tags": true
220
+ },
221
+ "peft": null,
222
+ "sample_weighting": null,
223
+ "rename_map": {},
224
+ "checkpoint_path": null
225
+ }