Hugging Face's logo

Sakits
/
pi05_async4_pickup_ball_v3

Safetensors
Model card Files
xet
 Community
Sakits commited on
Commit
4dddbf4
·
verified ·
1 Parent(s): ff3f13e

Upload folder using huggingface_hub

Browse files
Files changed (3) hide show
  1. config.json +88 -0
  2. model.safetensors +3 -0
  3. train_config.json +240 -0
config.json ADDED
@@ -0,0 +1,88 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "pi05",
3
+ "n_obs_steps": 1,
4
+ "input_features": {
5
+ "observation.state": {
6
+ "type": "STATE",
7
+ "shape": [
8
+ 6
9
+ ]
10
+ },
11
+ "observation.images.wrist": {
12
+ "type": "VISUAL",
13
+ "shape": [
14
+ 3,
15
+ 480,
16
+ 640
17
+ ]
18
+ },
19
+ "observation.images.top": {
20
+ "type": "VISUAL",
21
+ "shape": [
22
+ 3,
23
+ 720,
24
+ 1280
25
+ ]
26
+ }
27
+ },
28
+ "output_features": {
29
+ "action": {
30
+ "type": "ACTION",
31
+ "shape": [
32
+ 6
33
+ ]
34
+ }
35
+ },
36
+ "device": "cuda",
37
+ "use_amp": false,
38
+ "push_to_hub": false,
39
+ "repo_id": null,
40
+ "private": null,
41
+ "tags": null,
42
+ "license": null,
43
+ "pretrained_path": "lerobot/pi05_base",
44
+ "paligemma_variant": "gemma_2b",
45
+ "action_expert_variant": "gemma_300m",
46
+ "dtype": "bfloat16",
47
+ "chunk_size": 50,
48
+ "n_action_steps": 50,
49
+ "max_state_dim": 32,
50
+ "max_action_dim": 32,
51
+ "state_cond": true,
52
+ "num_inference_steps": 10,
53
+ "time_sampling_beta_alpha": 1.5,
54
+ "time_sampling_beta_beta": 1.0,
55
+ "time_sampling_scale": 0.999,
56
+ "time_sampling_offset": 0.001,
57
+ "min_period": 0.004,
58
+ "max_period": 4.0,
59
+ "image_resolution": [
60
+ 224,
61
+ 224
62
+ ],
63
+ "empty_cameras": 0,
64
+ "tokenizer_max_length": 200,
65
+ "normalization_mapping": {
66
+ "VISUAL": "IDENTITY",
67
+ "STATE": "MEAN_STD",
68
+ "ACTION": "MEAN_STD"
69
+ },
70
+ "gradient_checkpointing": false,
71
+ "compile_model": false,
72
+ "compile_mode": "max-autotune",
73
+ "fuse_qkv": false,
74
+ "fuse_gate_up": false,
75
+ "optimizer_lr": 2.5e-05,
76
+ "optimizer_betas": [
77
+ 0.9,
78
+ 0.95
79
+ ],
80
+ "optimizer_eps": 1e-08,
81
+ "optimizer_weight_decay": 0.01,
82
+ "optimizer_grad_clip_norm": 1.0,
83
+ "scheduler_warmup_steps": 1000,
84
+ "scheduler_decay_steps": 30000,
85
+ "scheduler_decay_lr": 2.5e-06,
86
+ "vlm_config": {},
87
+ "action_expert_config": {}
88
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e60f094c971b043b27f6f98f126ec376187773d478702b6d018558ed2fef574e
3
+ size 7481485688
train_config.json ADDED
@@ -0,0 +1,240 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "Sakits/so101_pickup_ball_v3",
4
+ "root": null,
5
+ "episodes": null,
6
+ "image_transforms": {
7
+ "enable": false,
8
+ "max_num_transforms": 3,
9
+ "random_order": false,
10
+ "tfs": {
11
+ "brightness": {
12
+ "weight": 1.0,
13
+ "type": "ColorJitter",
14
+ "kwargs": {
15
+ "brightness": [
16
+ 0.8,
17
+ 1.2
18
+ ]
19
+ }
20
+ },
21
+ "contrast": {
22
+ "weight": 1.0,
23
+ "type": "ColorJitter",
24
+ "kwargs": {
25
+ "contrast": [
26
+ 0.8,
27
+ 1.2
28
+ ]
29
+ }
30
+ },
31
+ "saturation": {
32
+ "weight": 1.0,
33
+ "type": "ColorJitter",
34
+ "kwargs": {
35
+ "saturation": [
36
+ 0.5,
37
+ 1.5
38
+ ]
39
+ }
40
+ },
41
+ "hue": {
42
+ "weight": 1.0,
43
+ "type": "ColorJitter",
44
+ "kwargs": {
45
+ "hue": [
46
+ -0.05,
47
+ 0.05
48
+ ]
49
+ }
50
+ },
51
+ "sharpness": {
52
+ "weight": 1.0,
53
+ "type": "SharpnessJitter",
54
+ "kwargs": {
55
+ "sharpness": [
56
+ 0.5,
57
+ 1.5
58
+ ]
59
+ }
60
+ },
61
+ "affine": {
62
+ "weight": 1.0,
63
+ "type": "RandomAffine",
64
+ "kwargs": {
65
+ "degrees": [
66
+ -5.0,
67
+ 5.0
68
+ ],
69
+ "translate": [
70
+ 0.05,
71
+ 0.05
72
+ ]
73
+ }
74
+ }
75
+ }
76
+ },
77
+ "revision": null,
78
+ "use_imagenet_stats": true,
79
+ "video_backend": "torchcodec",
80
+ "streaming": false
81
+ },
82
+ "env": null,
83
+ "policy": {
84
+ "type": "pi05",
85
+ "n_obs_steps": 1,
86
+ "input_features": {
87
+ "observation.state": {
88
+ "type": "STATE",
89
+ "shape": [
90
+ 6
91
+ ]
92
+ },
93
+ "observation.images.wrist": {
94
+ "type": "VISUAL",
95
+ "shape": [
96
+ 3,
97
+ 480,
98
+ 640
99
+ ]
100
+ },
101
+ "observation.images.top": {
102
+ "type": "VISUAL",
103
+ "shape": [
104
+ 3,
105
+ 720,
106
+ 1280
107
+ ]
108
+ }
109
+ },
110
+ "output_features": {
111
+ "action": {
112
+ "type": "ACTION",
113
+ "shape": [
114
+ 6
115
+ ]
116
+ }
117
+ },
118
+ "device": "cuda",
119
+ "use_amp": false,
120
+ "push_to_hub": false,
121
+ "repo_id": null,
122
+ "private": null,
123
+ "tags": null,
124
+ "license": null,
125
+ "pretrained_path": "lerobot/pi05_base",
126
+ "paligemma_variant": "gemma_2b",
127
+ "action_expert_variant": "gemma_300m",
128
+ "dtype": "bfloat16",
129
+ "chunk_size": 50,
130
+ "n_action_steps": 50,
131
+ "max_state_dim": 32,
132
+ "max_action_dim": 32,
133
+ "state_cond": true,
134
+ "num_inference_steps": 10,
135
+ "time_sampling_beta_alpha": 1.5,
136
+ "time_sampling_beta_beta": 1.0,
137
+ "time_sampling_scale": 0.999,
138
+ "time_sampling_offset": 0.001,
139
+ "min_period": 0.004,
140
+ "max_period": 4.0,
141
+ "image_resolution": [
142
+ 224,
143
+ 224
144
+ ],
145
+ "empty_cameras": 0,
146
+ "tokenizer_max_length": 200,
147
+ "normalization_mapping": {
148
+ "VISUAL": "IDENTITY",
149
+ "STATE": "MEAN_STD",
150
+ "ACTION": "MEAN_STD"
151
+ },
152
+ "gradient_checkpointing": false,
153
+ "compile_model": false,
154
+ "compile_mode": "max-autotune",
155
+ "fuse_qkv": false,
156
+ "fuse_gate_up": false,
157
+ "optimizer_lr": 2.5e-05,
158
+ "optimizer_betas": [
159
+ 0.9,
160
+ 0.95
161
+ ],
162
+ "optimizer_eps": 1e-08,
163
+ "optimizer_weight_decay": 0.01,
164
+ "optimizer_grad_clip_norm": 1.0,
165
+ "scheduler_warmup_steps": 1000,
166
+ "scheduler_decay_steps": 30000,
167
+ "scheduler_decay_lr": 2.5e-06,
168
+ "vlm_config": {},
169
+ "action_expert_config": {}
170
+ },
171
+ "output_dir": "outputs/train/pi05_async4_pickup_ball_v3",
172
+ "job_name": "pi05_async4_pickup_ball_v3",
173
+ "resume": false,
174
+ "seed": 1000,
175
+ "num_workers": 4,
176
+ "batch_size": 8,
177
+ "steps": 50000,
178
+ "eval_freq": 20000,
179
+ "log_freq": 200,
180
+ "save_checkpoint": true,
181
+ "save_freq": 10000,
182
+ "use_policy_training_preset": false,
183
+ "optimizer": {
184
+ "type": "adamw",
185
+ "lr": 5e-05,
186
+ "weight_decay": 1e-10,
187
+ "grad_clip_norm": 10.0,
188
+ "betas": [
189
+ 0.9,
190
+ 0.95
191
+ ],
192
+ "eps": 1e-08
193
+ },
194
+ "scheduler": {
195
+ "type": "cosine_decay_with_warmup",
196
+ "num_warmup_steps": 1000,
197
+ "num_decay_steps": 50000,
198
+ "peak_lr": 5e-05,
199
+ "decay_lr": 2.5e-06
200
+ },
201
+ "eval": {
202
+ "n_episodes": 50,
203
+ "batch_size": 50,
204
+ "use_async_envs": false
205
+ },
206
+ "wandb": {
207
+ "enable": true,
208
+ "disable_artifact": true,
209
+ "project": "vlash",
210
+ "entity": null,
211
+ "notes": null,
212
+ "run_id": "9hlvmvt2",
213
+ "mode": null
214
+ },
215
+ "checkpoint_path": null,
216
+ "rename_map": {},
217
+ "max_delay_steps": 4,
218
+ "grad_accum_steps": 1,
219
+ "shared_observation": true,
220
+ "lora": {
221
+ "enable": false,
222
+ "backend": "peft",
223
+ "r": 16,
224
+ "alpha": 16,
225
+ "dropout": 0.0,
226
+ "extra_trainable_modules": [],
227
+ "target_modules": [
228
+ "q_proj",
229
+ "k_proj",
230
+ "v_proj",
231
+ "o_proj",
232
+ "up_proj",
233
+ "down_proj",
234
+ "gate_proj"
235
+ ],
236
+ "use_qlora": false,
237
+ "qlora_quant_type": "nf4",
238
+ "qlora_compute_dtype": "bfloat16"
239
+ }
240
+ }