Hugging Face's logo

imstevenpmwork
/
xvla-tshirt-train-folding

Robotics
LeRobot
Safetensors
xvla
Model card Files
xet
 Community
imstevenpmwork HF Staff commited on
Commit
771b40f
·
verified ·
1 Parent(s): c4bf950

Upload policy weights, train config and readme

Browse files
Files changed (4) hide show
  1. README.md +62 -0
  2. config.json +195 -0
  3. model.safetensors +3 -0
  4. train_config.json +331 -0
README.md ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ datasets: lerobot-data-collection/folding_2025-12-06_level1
3
+ library_name: lerobot
4
+ license: apache-2.0
5
+ model_name: xvla
6
+ pipeline_tag: robotics
7
+ tags:
8
+ - xvla
9
+ - robotics
10
+ - lerobot
11
+ ---
12
+
13
+ # Model Card for xvla
14
+
15
+ <!-- Provide a quick summary of what the model is/does. -->
16
+
17
+
18
+ _Model type not recognized — please update this template._
19
+
20
+
21
+ This policy has been trained and pushed to the Hub using [LeRobot](https://github.com/huggingface/lerobot).
22
+ See the full documentation at [LeRobot Docs](https://huggingface.co/docs/lerobot/index).
23
+
24
+ ---
25
+
26
+ ## How to Get Started with the Model
27
+
28
+ For a complete walkthrough, see the [training guide](https://huggingface.co/docs/lerobot/il_robots#train-a-policy).
29
+ Below is the short version on how to train and run inference/eval:
30
+
31
+ ### Train from scratch
32
+
33
+ ```bash
34
+ lerobot-train \
35
+ --dataset.repo_id=${HF_USER}/<dataset> \
36
+ --policy.type=act \
37
+ --output_dir=outputs/train/<desired_policy_repo_id> \
38
+ --job_name=lerobot_training \
39
+ --policy.device=cuda \
40
+ --policy.repo_id=${HF_USER}/<desired_policy_repo_id>
41
+ --wandb.enable=true
42
+ ```
43
+
44
+ _Writes checkpoints to `outputs/train/<desired_policy_repo_id>/checkpoints/`._
45
+
46
+ ### Evaluate the policy/run inference
47
+
48
+ ```bash
49
+ lerobot-record \
50
+ --robot.type=so100_follower \
51
+ --dataset.repo_id=<hf_user>/eval_<dataset> \
52
+ --policy.path=<hf_user>/<desired_policy_repo_id> \
53
+ --episodes=10
54
+ ```
55
+
56
+ Prefix the dataset repo with **eval\_** and supply `--policy.path` pointing to a local or hub checkpoint.
57
+
58
+ ---
59
+
60
+ ## Model Details
61
+
62
+ - **License:** apache-2.0
config.json ADDED
@@ -0,0 +1,195 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "xvla",
3
+ "n_obs_steps": 1,
4
+ "input_features": {
5
+ "observation.images.image": {
6
+ "type": "VISUAL",
7
+ "shape": [
8
+ 3,
9
+ 256,
10
+ 256
11
+ ]
12
+ },
13
+ "observation.images.image2": {
14
+ "type": "VISUAL",
15
+ "shape": [
16
+ 3,
17
+ 256,
18
+ 256
19
+ ]
20
+ },
21
+ "observation.state": {
22
+ "type": "STATE",
23
+ "shape": [
24
+ 8
25
+ ]
26
+ },
27
+ "observation.images.empty_camera_0": {
28
+ "type": "VISUAL",
29
+ "shape": [
30
+ 3,
31
+ 224,
32
+ 224
33
+ ]
34
+ }
35
+ },
36
+ "output_features": {
37
+ "action": {
38
+ "type": "ACTION",
39
+ "shape": [
40
+ 16
41
+ ]
42
+ }
43
+ },
44
+ "device": "cuda",
45
+ "use_amp": false,
46
+ "push_to_hub": true,
47
+ "repo_id": "imstevenpmwork/xvla-tshirt-train-folding",
48
+ "private": null,
49
+ "tags": null,
50
+ "license": null,
51
+ "pretrained_path": "lerobot/xvla-folding",
52
+ "chunk_size": 30,
53
+ "n_action_steps": 30,
54
+ "dtype": "bfloat16",
55
+ "normalization_mapping": {
56
+ "STATE": "IDENTITY",
57
+ "ACTION": "MEAN_STD",
58
+ "VISUAL": "IDENTITY"
59
+ },
60
+ "florence_config": {
61
+ "model_type": "florence2",
62
+ "bos_token_id": 0,
63
+ "eos_token_id": 2,
64
+ "ignore_index": -100,
65
+ "pad_token_id": 1,
66
+ "projection_dim": 1024,
67
+ "text_config": {
68
+ "vocab_size": 51289,
69
+ "activation_dropout": 0.1,
70
+ "activation_function": "gelu",
71
+ "attention_dropout": 0.1,
72
+ "d_model": 1024,
73
+ "decoder_attention_heads": 16,
74
+ "decoder_layers": 12,
75
+ "encoder_attention_heads": 16,
76
+ "encoder_layers": 12,
77
+ "dropout": 0.1,
78
+ "max_position_embeddings": 4096,
79
+ "num_hidden_layers": 12,
80
+ "num_beams": 3
81
+ },
82
+ "vision_config": {
83
+ "model_type": "davit",
84
+ "drop_path_rate": 0.1,
85
+ "patch_size": [
86
+ 7,
87
+ 3,
88
+ 3,
89
+ 3
90
+ ],
91
+ "patch_stride": [
92
+ 4,
93
+ 2,
94
+ 2,
95
+ 2
96
+ ],
97
+ "patch_padding": [
98
+ 3,
99
+ 1,
100
+ 1,
101
+ 1
102
+ ],
103
+ "patch_prenorm": [
104
+ false,
105
+ true,
106
+ true,
107
+ true
108
+ ],
109
+ "enable_checkpoint": false,
110
+ "dim_embed": [
111
+ 256,
112
+ 512,
113
+ 1024,
114
+ 2048
115
+ ],
116
+ "num_heads": [
117
+ 8,
118
+ 16,
119
+ 32,
120
+ 64
121
+ ],
122
+ "num_groups": [
123
+ 8,
124
+ 16,
125
+ 32,
126
+ 64
127
+ ],
128
+ "depths": [
129
+ 1,
130
+ 1,
131
+ 9,
132
+ 1
133
+ ],
134
+ "window_size": 12,
135
+ "projection_dim": 1024,
136
+ "visual_temporal_embedding": {
137
+ "type": "COSINE",
138
+ "max_temporal_embeddings": 100
139
+ },
140
+ "image_pos_embed": {
141
+ "type": "learned_abs_2d",
142
+ "max_pos_embeddings": 50
143
+ },
144
+ "image_feature_source": [
145
+ "spatial_avg_pool",
146
+ "temporal_avg_pool"
147
+ ]
148
+ },
149
+ "vocab_size": 51289,
150
+ "torch_dtype": "float16",
151
+ "is_encoder_decoder": true
152
+ },
153
+ "tokenizer_name": "facebook/bart-large",
154
+ "tokenizer_max_length": 1024,
155
+ "tokenizer_padding_side": "right",
156
+ "pad_language_to": "max_length",
157
+ "hidden_size": 1024,
158
+ "depth": 24,
159
+ "num_heads": 16,
160
+ "mlp_ratio": 4.0,
161
+ "num_domains": 30,
162
+ "len_soft_prompts": 32,
163
+ "dim_time": 32,
164
+ "max_len_seq": 512,
165
+ "use_hetero_proj": false,
166
+ "action_mode": "auto",
167
+ "num_denoising_steps": 10,
168
+ "use_proprio": true,
169
+ "max_state_dim": 20,
170
+ "max_action_dim": 20,
171
+ "domain_feature_key": null,
172
+ "resize_imgs_with_padding": [
173
+ 224,
174
+ 224
175
+ ],
176
+ "num_image_views": 3,
177
+ "empty_cameras": 0,
178
+ "freeze_vision_encoder": false,
179
+ "freeze_language_encoder": false,
180
+ "train_policy_transformer": true,
181
+ "train_soft_prompts": true,
182
+ "optimizer_lr": 0.0001,
183
+ "optimizer_betas": [
184
+ 0.9,
185
+ 0.95
186
+ ],
187
+ "optimizer_eps": 1e-08,
188
+ "optimizer_weight_decay": 0.0001,
189
+ "optimizer_grad_clip_norm": 10.0,
190
+ "optimizer_soft_prompt_lr_scale": 1.0,
191
+ "optimizer_soft_prompt_warmup_lr_scale": null,
192
+ "scheduler_warmup_steps": 1000,
193
+ "scheduler_decay_steps": 30000,
194
+ "scheduler_decay_lr": 2.5e-06
195
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6716f00ed7c465557b2ecacd15cd6b577eff6f5343036c0d908737bdea80cf5
3
+ size 1759596986
train_config.json ADDED
@@ -0,0 +1,331 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "lerobot-data-collection/folding_2025-12-06_level1",
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": "xvla",
85
+ "n_obs_steps": 1,
86
+ "input_features": {
87
+ "observation.images.image": {
88
+ "type": "VISUAL",
89
+ "shape": [
90
+ 3,
91
+ 256,
92
+ 256
93
+ ]
94
+ },
95
+ "observation.images.image2": {
96
+ "type": "VISUAL",
97
+ "shape": [
98
+ 3,
99
+ 256,
100
+ 256
101
+ ]
102
+ },
103
+ "observation.state": {
104
+ "type": "STATE",
105
+ "shape": [
106
+ 8
107
+ ]
108
+ },
109
+ "observation.images.empty_camera_0": {
110
+ "type": "VISUAL",
111
+ "shape": [
112
+ 3,
113
+ 224,
114
+ 224
115
+ ]
116
+ }
117
+ },
118
+ "output_features": {
119
+ "action": {
120
+ "type": "ACTION",
121
+ "shape": [
122
+ 16
123
+ ]
124
+ }
125
+ },
126
+ "device": "cuda",
127
+ "use_amp": false,
128
+ "push_to_hub": true,
129
+ "repo_id": "imstevenpmwork/xvla-tshirt-train-folding",
130
+ "private": null,
131
+ "tags": null,
132
+ "license": null,
133
+ "pretrained_path": "lerobot/xvla-folding",
134
+ "chunk_size": 30,
135
+ "n_action_steps": 30,
136
+ "dtype": "bfloat16",
137
+ "normalization_mapping": {
138
+ "STATE": "IDENTITY",
139
+ "ACTION": "MEAN_STD",
140
+ "VISUAL": "IDENTITY"
141
+ },
142
+ "florence_config": {
143
+ "model_type": "florence2",
144
+ "bos_token_id": 0,
145
+ "eos_token_id": 2,
146
+ "ignore_index": -100,
147
+ "pad_token_id": 1,
148
+ "projection_dim": 1024,
149
+ "text_config": {
150
+ "vocab_size": 51289,
151
+ "activation_dropout": 0.1,
152
+ "activation_function": "gelu",
153
+ "attention_dropout": 0.1,
154
+ "d_model": 1024,
155
+ "decoder_attention_heads": 16,
156
+ "decoder_layers": 12,
157
+ "encoder_attention_heads": 16,
158
+ "encoder_layers": 12,
159
+ "dropout": 0.1,
160
+ "max_position_embeddings": 4096,
161
+ "num_hidden_layers": 12,
162
+ "num_beams": 3
163
+ },
164
+ "vision_config": {
165
+ "model_type": "davit",
166
+ "drop_path_rate": 0.1,
167
+ "patch_size": [
168
+ 7,
169
+ 3,
170
+ 3,
171
+ 3
172
+ ],
173
+ "patch_stride": [
174
+ 4,
175
+ 2,
176
+ 2,
177
+ 2
178
+ ],
179
+ "patch_padding": [
180
+ 3,
181
+ 1,
182
+ 1,
183
+ 1
184
+ ],
185
+ "patch_prenorm": [
186
+ false,
187
+ true,
188
+ true,
189
+ true
190
+ ],
191
+ "enable_checkpoint": false,
192
+ "dim_embed": [
193
+ 256,
194
+ 512,
195
+ 1024,
196
+ 2048
197
+ ],
198
+ "num_heads": [
199
+ 8,
200
+ 16,
201
+ 32,
202
+ 64
203
+ ],
204
+ "num_groups": [
205
+ 8,
206
+ 16,
207
+ 32,
208
+ 64
209
+ ],
210
+ "depths": [
211
+ 1,
212
+ 1,
213
+ 9,
214
+ 1
215
+ ],
216
+ "window_size": 12,
217
+ "projection_dim": 1024,
218
+ "visual_temporal_embedding": {
219
+ "type": "COSINE",
220
+ "max_temporal_embeddings": 100
221
+ },
222
+ "image_pos_embed": {
223
+ "type": "learned_abs_2d",
224
+ "max_pos_embeddings": 50
225
+ },
226
+ "image_feature_source": [
227
+ "spatial_avg_pool",
228
+ "temporal_avg_pool"
229
+ ]
230
+ },
231
+ "vocab_size": 51289,
232
+ "torch_dtype": "float16",
233
+ "is_encoder_decoder": true
234
+ },
235
+ "tokenizer_name": "facebook/bart-large",
236
+ "tokenizer_max_length": 1024,
237
+ "tokenizer_padding_side": "right",
238
+ "pad_language_to": "max_length",
239
+ "hidden_size": 1024,
240
+ "depth": 24,
241
+ "num_heads": 16,
242
+ "mlp_ratio": 4.0,
243
+ "num_domains": 30,
244
+ "len_soft_prompts": 32,
245
+ "dim_time": 32,
246
+ "max_len_seq": 512,
247
+ "use_hetero_proj": false,
248
+ "action_mode": "auto",
249
+ "num_denoising_steps": 10,
250
+ "use_proprio": true,
251
+ "max_state_dim": 20,
252
+ "max_action_dim": 20,
253
+ "domain_feature_key": null,
254
+ "resize_imgs_with_padding": [
255
+ 224,
256
+ 224
257
+ ],
258
+ "num_image_views": 3,
259
+ "empty_cameras": 0,
260
+ "freeze_vision_encoder": false,
261
+ "freeze_language_encoder": false,
262
+ "train_policy_transformer": true,
263
+ "train_soft_prompts": true,
264
+ "optimizer_lr": 0.0001,
265
+ "optimizer_betas": [
266
+ 0.9,
267
+ 0.95
268
+ ],
269
+ "optimizer_eps": 1e-08,
270
+ "optimizer_weight_decay": 0.0001,
271
+ "optimizer_grad_clip_norm": 10.0,
272
+ "optimizer_soft_prompt_lr_scale": 1.0,
273
+ "optimizer_soft_prompt_warmup_lr_scale": null,
274
+ "scheduler_warmup_steps": 1000,
275
+ "scheduler_decay_steps": 30000,
276
+ "scheduler_decay_lr": 2.5e-06
277
+ },
278
+ "output_dir": "outputs/train/2025-12-16/10-20-18_xvla-tshirt-train-folding",
279
+ "job_name": "xvla-tshirt-train-folding",
280
+ "resume": false,
281
+ "seed": 1000,
282
+ "num_workers": 4,
283
+ "batch_size": 32,
284
+ "steps": 30000,
285
+ "eval_freq": 20000,
286
+ "log_freq": 30,
287
+ "tolerance_s": 0.01,
288
+ "save_checkpoint": true,
289
+ "save_freq": 10000,
290
+ "use_policy_training_preset": true,
291
+ "optimizer": {
292
+ "type": "xvla-adamw",
293
+ "lr": 0.0001,
294
+ "weight_decay": 0.0001,
295
+ "grad_clip_norm": 10.0,
296
+ "betas": [
297
+ 0.9,
298
+ 0.95
299
+ ],
300
+ "eps": 1e-08,
301
+ "soft_prompt_lr_scale": 1.0,
302
+ "soft_prompt_warmup_lr_scale": null
303
+ },
304
+ "scheduler": {
305
+ "type": "cosine_decay_with_warmup",
306
+ "num_warmup_steps": 1000,
307
+ "num_decay_steps": 30000,
308
+ "peak_lr": 0.0001,
309
+ "decay_lr": 2.5e-06
310
+ },
311
+ "eval": {
312
+ "n_episodes": 50,
313
+ "batch_size": 50,
314
+ "use_async_envs": false
315
+ },
316
+ "wandb": {
317
+ "enable": true,
318
+ "disable_artifact": true,
319
+ "project": "lerobot",
320
+ "entity": null,
321
+ "notes": null,
322
+ "run_id": "j8uzxwez",
323
+ "mode": null
324
+ },
325
+ "checkpoint_path": null,
326
+ "rename_map": {
327
+ "observation.images.base": "observation.images.empty_camera_0",
328
+ "observation.images.left_wrist": "observation.images.image",
329
+ "observation.images.right_wrist": "observation.images.image2"
330
+ }
331
+ }