11yiming commited on
Commit
4ea7b9a
·
verified ·
1 Parent(s): 8bfb1e5

Upload GR00T Eagle 3B RoboCasa GR1 24x30 fine-tuned checkpoint

Browse files
.gitattributes CHANGED
@@ -40,3 +40,4 @@ gr00t_qwen3vl_0.6b_libero/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs
40
  pi05_paligemma_libero_10_full_finetune_bs64/pi05_paligemma_libero_10_full_finetune_2026_05_15_09_15_10.jsonl filter=lfs diff=lfs merge=lfs -text
41
  pi0_paligemma_libero_10_full_finetune_bs64/pi0_paligemma_libero_10_full_finetune_2026_05_25_21_16_07.jsonl filter=lfs diff=lfs merge=lfs -text
42
  openvla_dino_siglip_llama2_libero_10_full_finetune_bs64/openvla_dino_siglip_llama2_libero_10_full_finetune_2026_05_27_09_35_11.jsonl filter=lfs diff=lfs merge=lfs -text
 
 
40
  pi05_paligemma_libero_10_full_finetune_bs64/pi05_paligemma_libero_10_full_finetune_2026_05_15_09_15_10.jsonl filter=lfs diff=lfs merge=lfs -text
41
  pi0_paligemma_libero_10_full_finetune_bs64/pi0_paligemma_libero_10_full_finetune_2026_05_25_21_16_07.jsonl filter=lfs diff=lfs merge=lfs -text
42
  openvla_dino_siglip_llama2_libero_10_full_finetune_bs64/openvla_dino_siglip_llama2_libero_10_full_finetune_2026_05_27_09_35_11.jsonl filter=lfs diff=lfs merge=lfs -text
43
+ gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug_2026_05_26_16_59_09.jsonl filter=lfs diff=lfs merge=lfs -text
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/checkpoints/step-010000-epoch-14-loss=0.0061.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5989eea1bba8bfdd77f1afc7f5bf60118fba126fe75cc384aec82fc39b86b85e
3
+ size 5448457056
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/config.json ADDED
@@ -0,0 +1,363 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_OFFICIAL_GR1_STATS_PATH": "work_dirs/official_groot_gr1_dataset_statistics.json",
3
+ "_STAT": "robocasa_gr1_24tasks_30ep",
4
+ "eval": {
5
+ "dataset": {
6
+ "transforms": [
7
+ {
8
+ "center_crop_scale": 0.95,
9
+ "embodiment_id": 24,
10
+ "img_key": "video.ego_view_bg_crop_pad_res256_freq20",
11
+ "normalize": false,
12
+ "resize_size": 224,
13
+ "type": "ProcessRobocasaEvalInputs"
14
+ },
15
+ {
16
+ "type": "RobocasaGR1N15Bridge"
17
+ },
18
+ {
19
+ "action_key": "action",
20
+ "norm_type": "min_max",
21
+ "normalize_states": false,
22
+ "state_dim": 64,
23
+ "state_key": "proprio",
24
+ "type": "NormalizeStatesAndActions"
25
+ },
26
+ {
27
+ "max_len": 900,
28
+ "num_images": 1,
29
+ "return_text": true,
30
+ "tokenizer": {
31
+ "model_path": "fluxvla/models/third_party_models/eagle2_hg_model",
32
+ "type": "PretrainedTokenizer"
33
+ },
34
+ "type": "ProcessPromptsWithImage"
35
+ },
36
+ {
37
+ "means": [
38
+ [
39
+ 127.5,
40
+ 127.5,
41
+ 127.5
42
+ ]
43
+ ],
44
+ "stds": [
45
+ [
46
+ 127.5,
47
+ 127.5,
48
+ 127.5
49
+ ]
50
+ ],
51
+ "type": "NormalizeImages"
52
+ }
53
+ ],
54
+ "type": "RobocasaEvalDataset",
55
+ "unnorm_key": "robocasa_gr1_24tasks_30ep"
56
+ },
57
+ "denormalize_action": {
58
+ "action_dim": 29,
59
+ "clip_actions": false,
60
+ "norm_type": "min_max",
61
+ "stats_order": "fluxvla",
62
+ "type": "DenormalizeRobocasaAction"
63
+ },
64
+ "eval_chunk_size": 16,
65
+ "max_episode_steps": 720,
66
+ "model_family": "groot",
67
+ "num_trials_per_task": 20,
68
+ "seed": 7,
69
+ "task_list": [
70
+ "gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env",
71
+ "gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env",
72
+ "gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env",
73
+ "gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env",
74
+ "gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env",
75
+ "gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env",
76
+ "gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env",
77
+ "gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env",
78
+ "gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env",
79
+ "gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env",
80
+ "gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env",
81
+ "gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env",
82
+ "gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env",
83
+ "gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env",
84
+ "gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env",
85
+ "gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env",
86
+ "gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env",
87
+ "gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env",
88
+ "gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env",
89
+ "gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env",
90
+ "gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env",
91
+ "gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env",
92
+ "gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env",
93
+ "gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env"
94
+ ],
95
+ "type": "RobocasaEvalRunner",
96
+ "unnorm_key": "robocasa_gr1_24tasks_30ep"
97
+ },
98
+ "inference_model": {
99
+ "pretrained_name_or_path": "./checkpoints/GR00T-N1.5-3B",
100
+ "type": "LlavaVLA",
101
+ "vla_head": {
102
+ "action_dim": 32,
103
+ "diffusion_model_cfg": {
104
+ "attention_head_dim": 48,
105
+ "cross_attention_dim": 2048,
106
+ "dropout": 0.2,
107
+ "final_dropout": true,
108
+ "interleave_self_attention": true,
109
+ "norm_type": "ada_norm",
110
+ "num_attention_heads": 32,
111
+ "num_layers": 16,
112
+ "output_dim": 1024,
113
+ "positional_embeddings": null
114
+ },
115
+ "hidden_size": 1024,
116
+ "input_embedding_dim": 1536,
117
+ "max_input_seq_len": 900,
118
+ "num_heads": 4,
119
+ "num_inference_timesteps": 4,
120
+ "num_layers": 1,
121
+ "num_steps": 16,
122
+ "ori_action_dim": 29,
123
+ "state_dim": 64,
124
+ "traj_length": 16,
125
+ "type": "FlowMatchingInferenceHead"
126
+ },
127
+ "vlm_backbone": {
128
+ "type": "EagleInferenceBackbone",
129
+ "vlm_config": {
130
+ "max_input_seq_len": 900
131
+ },
132
+ "vlm_path": "fluxvla/models/third_party_models/eagle2_hg_model"
133
+ }
134
+ },
135
+ "model": {
136
+ "freeze_projector": false,
137
+ "freeze_vlm_backbone": false,
138
+ "name_mapping": {
139
+ "vla_head": "action_head",
140
+ "vlm_backbone.vlm": "backbone.eagle_model"
141
+ },
142
+ "pretrained_name_or_path": "./checkpoints/GR00T-N1.5-3B",
143
+ "type": "LlavaVLA",
144
+ "vla_head": {
145
+ "action_dim": 32,
146
+ "hidden_size": 1024,
147
+ "input_embedding_dim": 1536,
148
+ "num_heads": 4,
149
+ "num_inference_timesteps": 4,
150
+ "num_layers": 1,
151
+ "num_steps": 16,
152
+ "ori_action_dim": 29,
153
+ "state_dim": 64,
154
+ "traj_length": 16,
155
+ "type": "FlowMatchingHead"
156
+ },
157
+ "vlm_backbone": {
158
+ "dtype": "bf16",
159
+ "tune_llm": false,
160
+ "tune_visual": true,
161
+ "type": "EagleBackbone",
162
+ "vlm_config": {
163
+ "max_input_seq_len": 900
164
+ },
165
+ "vlm_path": "fluxvla/models/third_party_models/eagle2_hg_model"
166
+ }
167
+ },
168
+ "runner": {
169
+ "change_key_name": false,
170
+ "collator": {
171
+ "keys": [
172
+ "states",
173
+ "observation.eepose",
174
+ "timestamp",
175
+ "images",
176
+ "img_masks",
177
+ "lang_tokens",
178
+ "lang_masks",
179
+ "actions",
180
+ "action_masks",
181
+ "embodiment_ids"
182
+ ],
183
+ "meta_keys": [
184
+ "task_description",
185
+ "prompt",
186
+ "info",
187
+ "stats"
188
+ ],
189
+ "type": "DictCollator"
190
+ },
191
+ "enable_gradient_checkpointing": false,
192
+ "enable_mixed_precision_training": true,
193
+ "learning_rate": 6e-05,
194
+ "lr_scheduler_type": "linear-warmup+cosine-decay",
195
+ "max_epochs": null,
196
+ "max_grad_norm": 1.0,
197
+ "max_keep_ckpts": 20,
198
+ "max_steps": 50000,
199
+ "metric": {
200
+ "active_trackers": [
201
+ "jsonl",
202
+ "wandb"
203
+ ],
204
+ "grad_accumulation_steps": 1,
205
+ "run_dir": "work_dirs",
206
+ "type": "VLAMetric",
207
+ "window_size": 1
208
+ },
209
+ "mixed_precision_dtype": "bf16",
210
+ "sampler": null,
211
+ "save_epoch_interval": 1,
212
+ "save_full_model": true,
213
+ "save_iter_interval": 5000,
214
+ "sharding_strategy": "no-shard",
215
+ "tokenizer": {
216
+ "model_path": "fluxvla/models/third_party_models/eagle2_hg_model",
217
+ "type": "PretrainedTokenizer"
218
+ },
219
+ "type": "FSDPTrainRunner",
220
+ "warmup_ratio": 0.05,
221
+ "weight_decay": 1e-05
222
+ },
223
+ "train_dataloader": {
224
+ "dataset": {
225
+ "dataset_statistics_path": "work_dirs/official_groot_gr1_dataset_statistics.json",
226
+ "datasets": [
227
+ {
228
+ "action_key": "action",
229
+ "action_window_size": 16,
230
+ "data_root_path": [
231
+ "./datasets/robocasa_fluxvla/PnPBottleToCabinetClose",
232
+ "./datasets/robocasa_fluxvla/PnPCanToDrawerClose",
233
+ "./datasets/robocasa_fluxvla/PnPCupToDrawerClose",
234
+ "./datasets/robocasa_fluxvla/PnPMilkToMicrowaveClose",
235
+ "./datasets/robocasa_fluxvla/PnPPotatoToMicrowaveClose",
236
+ "./datasets/robocasa_fluxvla/PnPWineToCabinetClose",
237
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToBasketSplitA",
238
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA",
239
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToPanSplitA",
240
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToPotSplitA",
241
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA",
242
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToBasketSplitA",
243
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToBowlSplitA",
244
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToPlateSplitA",
245
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToTieredshelfSplitA",
246
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToBowlSplitA",
247
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToCardboardboxSplitA",
248
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToPanSplitA",
249
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToPlateSplitA",
250
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToCardboardboxSplitA",
251
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToPlateSplitA",
252
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToPotSplitA",
253
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToTieredbasketSplitA",
254
+ "./datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToTieredshelfSplitA"
255
+ ],
256
+ "statistic_name": "robocasa_gr1_24tasks_30ep",
257
+ "transforms": [
258
+ {
259
+ "embodiment_id": 24,
260
+ "name_mappings": {
261
+ "actions": [
262
+ "actions"
263
+ ],
264
+ "observation.state": [
265
+ "states"
266
+ ]
267
+ },
268
+ "parquet_keys": [
269
+ "observation.state",
270
+ "timestamp",
271
+ "actions",
272
+ "info",
273
+ "stats",
274
+ "action_masks"
275
+ ],
276
+ "type": "ProcessParquetInputs",
277
+ "video_keys": [
278
+ "observation.images.ego_view"
279
+ ]
280
+ },
281
+ {
282
+ "type": "RobocasaGR1N15Bridge"
283
+ },
284
+ {
285
+ "type": "ParquetPrompter"
286
+ },
287
+ {
288
+ "max_len": 900,
289
+ "num_images": 1,
290
+ "tokenizer": {
291
+ "model_path": "fluxvla/models/third_party_models/eagle2_hg_model",
292
+ "type": "PretrainedTokenizer"
293
+ },
294
+ "type": "ProcessPromptsWithImage"
295
+ },
296
+ {
297
+ "scale": 0.95,
298
+ "type": "RandomCropImages"
299
+ },
300
+ {
301
+ "height": 224,
302
+ "type": "ResizeImages",
303
+ "width": 224
304
+ },
305
+ {
306
+ "brightness": 0.3,
307
+ "contrast": 0.4,
308
+ "hue": 0.08,
309
+ "saturation": 0.5,
310
+ "type": "ColorJitterImages"
311
+ },
312
+ {
313
+ "means": [
314
+ [
315
+ 127.5,
316
+ 127.5,
317
+ 127.5
318
+ ]
319
+ ],
320
+ "stds": [
321
+ [
322
+ 127.5,
323
+ 127.5,
324
+ 127.5
325
+ ]
326
+ ],
327
+ "type": "NormalizeImages"
328
+ },
329
+ {
330
+ "action_dim": 32,
331
+ "action_key": "action",
332
+ "norm_type": "min_max",
333
+ "normalize_states": false,
334
+ "state_dim": 64,
335
+ "state_key": "proprio",
336
+ "type": "NormalizeStatesAndActions"
337
+ }
338
+ ],
339
+ "type": "ParquetDataset",
340
+ "use_delta": false,
341
+ "window_start_idx": 0
342
+ }
343
+ ],
344
+ "name_mappings": {
345
+ "action": [
346
+ "action"
347
+ ],
348
+ "observation.state": [
349
+ "proprio"
350
+ ]
351
+ },
352
+ "statistic_keys": [
353
+ "observation.state",
354
+ "timestamp",
355
+ "action"
356
+ ],
357
+ "statistic_name": "robocasa_gr1_24tasks_30ep",
358
+ "type": "DistributedRepeatingDataset"
359
+ },
360
+ "per_device_batch_size": 16,
361
+ "per_device_num_workers": 4
362
+ }
363
+ }
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/config.yaml ADDED
@@ -0,0 +1,281 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ _OFFICIAL_GR1_STATS_PATH: work_dirs/official_groot_gr1_dataset_statistics.json
2
+ _STAT: robocasa_gr1_24tasks_30ep
3
+ eval:
4
+ dataset:
5
+ transforms:
6
+ - center_crop_scale: 0.95
7
+ embodiment_id: 24
8
+ img_key: video.ego_view_bg_crop_pad_res256_freq20
9
+ normalize: false
10
+ resize_size: 224
11
+ type: ProcessRobocasaEvalInputs
12
+ - type: RobocasaGR1N15Bridge
13
+ - action_key: action
14
+ norm_type: min_max
15
+ normalize_states: false
16
+ state_dim: 64
17
+ state_key: proprio
18
+ type: NormalizeStatesAndActions
19
+ - max_len: 900
20
+ num_images: 1
21
+ return_text: true
22
+ tokenizer:
23
+ model_path: fluxvla/models/third_party_models/eagle2_hg_model
24
+ type: PretrainedTokenizer
25
+ type: ProcessPromptsWithImage
26
+ - means:
27
+ - - 127.5
28
+ - 127.5
29
+ - 127.5
30
+ stds:
31
+ - - 127.5
32
+ - 127.5
33
+ - 127.5
34
+ type: NormalizeImages
35
+ type: RobocasaEvalDataset
36
+ unnorm_key: robocasa_gr1_24tasks_30ep
37
+ denormalize_action:
38
+ action_dim: 29
39
+ clip_actions: false
40
+ norm_type: min_max
41
+ stats_order: fluxvla
42
+ type: DenormalizeRobocasaAction
43
+ eval_chunk_size: 16
44
+ max_episode_steps: 720
45
+ model_family: groot
46
+ num_trials_per_task: 20
47
+ seed: 7
48
+ task_list:
49
+ - gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env
50
+ - gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env
51
+ - gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env
52
+ - gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env
53
+ - gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env
54
+ - gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env
55
+ - gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env
56
+ - gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env
57
+ - gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env
58
+ - gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env
59
+ - gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env
60
+ - gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env
61
+ - gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env
62
+ - gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env
63
+ - gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env
64
+ - gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env
65
+ - gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env
66
+ - gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env
67
+ - gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env
68
+ - gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env
69
+ - gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env
70
+ - gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env
71
+ - gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env
72
+ - gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env
73
+ type: RobocasaEvalRunner
74
+ unnorm_key: robocasa_gr1_24tasks_30ep
75
+ inference_model:
76
+ pretrained_name_or_path: ./checkpoints/GR00T-N1.5-3B
77
+ type: LlavaVLA
78
+ vla_head:
79
+ action_dim: 32
80
+ diffusion_model_cfg:
81
+ attention_head_dim: 48
82
+ cross_attention_dim: 2048
83
+ dropout: 0.2
84
+ final_dropout: true
85
+ interleave_self_attention: true
86
+ norm_type: ada_norm
87
+ num_attention_heads: 32
88
+ num_layers: 16
89
+ output_dim: 1024
90
+ positional_embeddings: null
91
+ hidden_size: 1024
92
+ input_embedding_dim: 1536
93
+ max_input_seq_len: 900
94
+ num_heads: 4
95
+ num_inference_timesteps: 4
96
+ num_layers: 1
97
+ num_steps: 16
98
+ ori_action_dim: 29
99
+ state_dim: 64
100
+ traj_length: 16
101
+ type: FlowMatchingInferenceHead
102
+ vlm_backbone:
103
+ type: EagleInferenceBackbone
104
+ vlm_config:
105
+ max_input_seq_len: 900
106
+ vlm_path: fluxvla/models/third_party_models/eagle2_hg_model
107
+ model:
108
+ freeze_projector: false
109
+ freeze_vlm_backbone: false
110
+ name_mapping:
111
+ vla_head: action_head
112
+ vlm_backbone.vlm: backbone.eagle_model
113
+ pretrained_name_or_path: ./checkpoints/GR00T-N1.5-3B
114
+ type: LlavaVLA
115
+ vla_head:
116
+ action_dim: 32
117
+ hidden_size: 1024
118
+ input_embedding_dim: 1536
119
+ num_heads: 4
120
+ num_inference_timesteps: 4
121
+ num_layers: 1
122
+ num_steps: 16
123
+ ori_action_dim: 29
124
+ state_dim: 64
125
+ traj_length: 16
126
+ type: FlowMatchingHead
127
+ vlm_backbone:
128
+ dtype: bf16
129
+ tune_llm: false
130
+ tune_visual: true
131
+ type: EagleBackbone
132
+ vlm_config:
133
+ max_input_seq_len: 900
134
+ vlm_path: fluxvla/models/third_party_models/eagle2_hg_model
135
+ runner:
136
+ change_key_name: false
137
+ collator:
138
+ keys:
139
+ - states
140
+ - observation.eepose
141
+ - timestamp
142
+ - images
143
+ - img_masks
144
+ - lang_tokens
145
+ - lang_masks
146
+ - actions
147
+ - action_masks
148
+ - embodiment_ids
149
+ meta_keys:
150
+ - task_description
151
+ - prompt
152
+ - info
153
+ - stats
154
+ type: DictCollator
155
+ enable_gradient_checkpointing: false
156
+ enable_mixed_precision_training: true
157
+ learning_rate: 6.0e-05
158
+ lr_scheduler_type: linear-warmup+cosine-decay
159
+ max_epochs: null
160
+ max_grad_norm: 1.0
161
+ max_keep_ckpts: 20
162
+ max_steps: 50000
163
+ metric:
164
+ active_trackers:
165
+ - jsonl
166
+ - wandb
167
+ grad_accumulation_steps: 1
168
+ run_dir: work_dirs
169
+ type: VLAMetric
170
+ window_size: 1
171
+ mixed_precision_dtype: bf16
172
+ sampler: null
173
+ save_epoch_interval: 1
174
+ save_full_model: true
175
+ save_iter_interval: 5000
176
+ sharding_strategy: no-shard
177
+ tokenizer:
178
+ model_path: fluxvla/models/third_party_models/eagle2_hg_model
179
+ type: PretrainedTokenizer
180
+ type: FSDPTrainRunner
181
+ warmup_ratio: 0.05
182
+ weight_decay: 1.0e-05
183
+ train_dataloader:
184
+ dataset:
185
+ dataset_statistics_path: work_dirs/official_groot_gr1_dataset_statistics.json
186
+ datasets:
187
+ - action_key: action
188
+ action_window_size: 16
189
+ data_root_path:
190
+ - ./datasets/robocasa_fluxvla/PnPBottleToCabinetClose
191
+ - ./datasets/robocasa_fluxvla/PnPCanToDrawerClose
192
+ - ./datasets/robocasa_fluxvla/PnPCupToDrawerClose
193
+ - ./datasets/robocasa_fluxvla/PnPMilkToMicrowaveClose
194
+ - ./datasets/robocasa_fluxvla/PnPPotatoToMicrowaveClose
195
+ - ./datasets/robocasa_fluxvla/PnPWineToCabinetClose
196
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToBasketSplitA
197
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA
198
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToPanSplitA
199
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToPotSplitA
200
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA
201
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToBasketSplitA
202
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToBowlSplitA
203
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToPlateSplitA
204
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToTieredshelfSplitA
205
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToBowlSplitA
206
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToCardboardboxSplitA
207
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToPanSplitA
208
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToPlateSplitA
209
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToCardboardboxSplitA
210
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToPlateSplitA
211
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToPotSplitA
212
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToTieredbasketSplitA
213
+ - ./datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToTieredshelfSplitA
214
+ statistic_name: robocasa_gr1_24tasks_30ep
215
+ transforms:
216
+ - embodiment_id: 24
217
+ name_mappings:
218
+ actions:
219
+ - actions
220
+ observation.state:
221
+ - states
222
+ parquet_keys:
223
+ - observation.state
224
+ - timestamp
225
+ - actions
226
+ - info
227
+ - stats
228
+ - action_masks
229
+ type: ProcessParquetInputs
230
+ video_keys:
231
+ - observation.images.ego_view
232
+ - type: RobocasaGR1N15Bridge
233
+ - type: ParquetPrompter
234
+ - max_len: 900
235
+ num_images: 1
236
+ tokenizer:
237
+ model_path: fluxvla/models/third_party_models/eagle2_hg_model
238
+ type: PretrainedTokenizer
239
+ type: ProcessPromptsWithImage
240
+ - scale: 0.95
241
+ type: RandomCropImages
242
+ - height: 224
243
+ type: ResizeImages
244
+ width: 224
245
+ - brightness: 0.3
246
+ contrast: 0.4
247
+ hue: 0.08
248
+ saturation: 0.5
249
+ type: ColorJitterImages
250
+ - means:
251
+ - - 127.5
252
+ - 127.5
253
+ - 127.5
254
+ stds:
255
+ - - 127.5
256
+ - 127.5
257
+ - 127.5
258
+ type: NormalizeImages
259
+ - action_dim: 32
260
+ action_key: action
261
+ norm_type: min_max
262
+ normalize_states: false
263
+ state_dim: 64
264
+ state_key: proprio
265
+ type: NormalizeStatesAndActions
266
+ type: ParquetDataset
267
+ use_delta: false
268
+ window_start_idx: 0
269
+ name_mappings:
270
+ action:
271
+ - action
272
+ observation.state:
273
+ - proprio
274
+ statistic_keys:
275
+ - observation.state
276
+ - timestamp
277
+ - action
278
+ statistic_name: robocasa_gr1_24tasks_30ep
279
+ type: DistributedRepeatingDataset
280
+ per_device_batch_size: 16
281
+ per_device_num_workers: 4
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/dataset_statistics.json ADDED
@@ -0,0 +1,396 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "robocasa_gr1_24tasks_30ep": {
3
+ "proprio": {
4
+ "mean": [
5
+ -0.1721954196691514,
6
+ 0.18329788744449615,
7
+ -0.1360361576080323,
8
+ -1.0807313919067383,
9
+ 0.14450088143348713,
10
+ 0.13195440173149092,
11
+ -0.08306947350502007,
12
+ 0.21940384805202476,
13
+ 0.21819256246089933,
14
+ 0.20426052808761605,
15
+ 0.22833923995494854,
16
+ 0.06841608136892328,
17
+ 0.9091449975967411,
18
+ -0.1590277999639513,
19
+ -0.2511773109436038,
20
+ 0.08599518239498137,
21
+ -1.1420482397079474,
22
+ 0.21259814500808735,
23
+ 0.04237671568989762,
24
+ 0.03427042067050932,
25
+ 0.34993353486061096,
26
+ 0.3412855267524723,
27
+ 0.3271332383155817,
28
+ 0.3575766980648041,
29
+ 0.09723818302154554,
30
+ 1.0641368627548227,
31
+ 0.014510802924633035,
32
+ 0.014398206025362044,
33
+ -0.00020241182937752458
34
+ ],
35
+ "std": [
36
+ 0.4068214595317843,
37
+ 0.18259239196777355,
38
+ 0.2513278424739838,
39
+ 0.5786943435668955,
40
+ 0.33050763607025135,
41
+ 0.28609949350357067,
42
+ 0.3013432323932649,
43
+ 0.42784574627876293,
44
+ 0.4130409955978396,
45
+ 0.3710329830646513,
46
+ 0.4229725301265714,
47
+ 0.18603332340717335,
48
+ 0.7297042012214684,
49
+ 0.5631562471389767,
50
+ 0.24450603127479503,
51
+ 0.3282549381256103,
52
+ 0.7310536503791792,
53
+ 0.48139148950576843,
54
+ 0.35611921548843384,
55
+ 0.5090531110763546,
56
+ 0.49147421121597257,
57
+ 0.46545258164405845,
58
+ 0.44516301155090443,
59
+ 0.4866269826889039,
60
+ 0.1814946085214615,
61
+ 0.5735033154487607,
62
+ 0.1227026358246804,
63
+ 0.03244836255908008,
64
+ 0.007282953243702651
65
+ ],
66
+ "min": [
67
+ -1.7593636512756348,
68
+ -0.08899051696062088,
69
+ -1.6571800708770752,
70
+ -2.516343355178833,
71
+ -2.307133436203003,
72
+ -1.4132264852523804,
73
+ -1.4408576488494873,
74
+ -3.262805700302124,
75
+ -2.7771592140197754,
76
+ -3.4712820053100586,
77
+ -2.7828404903411865,
78
+ -2.197380781173706,
79
+ -2.636387825012207,
80
+ -2.2743337154388428,
81
+ -2.999122142791748,
82
+ -1.7424250841140747,
83
+ -2.536545753479004,
84
+ -2.6034910678863525,
85
+ -1.5032434463500977,
86
+ -1.4913568496704102,
87
+ -1.7595213651657104,
88
+ -1.8982670307159424,
89
+ -1.8622381687164307,
90
+ -1.8314869403839111,
91
+ -1.0045679807662964,
92
+ -0.31411829590797424,
93
+ -0.8131351470947266,
94
+ -0.4901888370513916,
95
+ -0.7304351925849915
96
+ ],
97
+ "max": [
98
+ 1.3570959568023682,
99
+ 1.5781080722808838,
100
+ 1.7425614595413208,
101
+ 0.04977008327841759,
102
+ 2.521491289138794,
103
+ 1.5104753971099854,
104
+ 1.3451268672943115,
105
+ 2.8765757083892822,
106
+ 2.6975045204162598,
107
+ 3.3291828632354736,
108
+ 2.7148845195770264,
109
+ 3.664421558380127,
110
+ 3.5466625690460205,
111
+ 1.7399436235427856,
112
+ 0.07876992970705032,
113
+ 2.5426626205444336,
114
+ 0.048908114433288574,
115
+ 3.000918388366699,
116
+ 1.4513440132141113,
117
+ 1.5000383853912354,
118
+ 2.397510290145874,
119
+ 2.4422738552093506,
120
+ 3.0158462524414062,
121
+ 2.457378625869751,
122
+ 1.4605257511138916,
123
+ 2.168534278869629,
124
+ 1.0323854684829712,
125
+ 0.7102982997894287,
126
+ 0.43762317299842834
127
+ ],
128
+ "q01": [
129
+ -1.6011584222316746,
130
+ -0.00048318122862838164,
131
+ -0.9488833963871,
132
+ -2.4362342357635502,
133
+ -0.9403592050075524,
134
+ -0.6489510446786882,
135
+ -0.9641011685132993,
136
+ -0.17130047902464876,
137
+ -0.0025680272374302166,
138
+ -0.0029985445667989555,
139
+ -0.002490740863140675,
140
+ -0.14894675761461273,
141
+ -0.00027310136792948466,
142
+ -1.4233640372753156,
143
+ -1.0436977982521036,
144
+ -0.7169023394584662,
145
+ -2.4936094284057644,
146
+ -0.7377233326435084,
147
+ -0.8036172896623616,
148
+ -0.9991738587617871,
149
+ -0.004218762856908145,
150
+ -0.004575904295779761,
151
+ -0.005576773989014321,
152
+ -0.0047070790780708175,
153
+ -0.18967307358980165,
154
+ 0.03200087323784828,
155
+ -0.3365072563290599,
156
+ -0.02904987111687659,
157
+ -0.02606033049523827
158
+ ],
159
+ "q99": [
160
+ 0.6305126994848259,
161
+ 0.7966297268867486,
162
+ 0.3873166218399999,
163
+ -0.07399308793246723,
164
+ 1.0455405175685892,
165
+ 0.8574009954929348,
166
+ 0.4838866069912913,
167
+ 1.5014824867248553,
168
+ 1.501653671264647,
169
+ 1.4043268024921445,
170
+ 1.5127789616584784,
171
+ 0.781709930300714,
172
+ 1.872669064998624,
173
+ 1.167767405509949,
174
+ 0.0006303858070168673,
175
+ 0.8771779745817198,
176
+ -0.0582109544426202,
177
+ 1.470731711387636,
178
+ 0.8474899172782918,
179
+ 1.2271583080291764,
180
+ 1.4994217157363876,
181
+ 1.5029035806655904,
182
+ 1.6614551365375507,
183
+ 1.5188306391239186,
184
+ 0.6953653633594535,
185
+ 1.8182492077350652,
186
+ 0.48663658648729347,
187
+ 0.12612193301320107,
188
+ 0.022280203178524964
189
+ ]
190
+ },
191
+ "timestamp": {
192
+ "mean": [
193
+ 0.0
194
+ ],
195
+ "std": [
196
+ 1.0
197
+ ],
198
+ "min": [
199
+ 0.0
200
+ ],
201
+ "max": [
202
+ 0.0
203
+ ],
204
+ "q01": null,
205
+ "q99": null
206
+ },
207
+ "action": {
208
+ "mean": [
209
+ -0.17771673202514657,
210
+ 0.18351773917675007,
211
+ -0.13720993697643286,
212
+ -1.0842658281326296,
213
+ 0.14675945043563873,
214
+ 0.12921456992626204,
215
+ -0.09504522383213035,
216
+ -0.22721463441848752,
217
+ -0.21956320106983204,
218
+ -0.2144497036933901,
219
+ -0.21446840465068817,
220
+ -0.48631641268730147,
221
+ 1.6838233470916744,
222
+ -0.16953834891319267,
223
+ -0.2535267770290373,
224
+ 0.082216002047062,
225
+ -1.1492931842803968,
226
+ 0.2176141738891601,
227
+ 0.04310995712876314,
228
+ 0.04133718833327294,
229
+ -0.543570637702943,
230
+ -0.5304322242736823,
231
+ -0.5179098844528184,
232
+ -0.4968528747558592,
233
+ -1.1343770027160642,
234
+ 2.128209590911863,
235
+ 0.014836843125522158,
236
+ 0.010428135283291354,
237
+ -0.00014937532250769447
238
+ ],
239
+ "std": [
240
+ 0.4174487590789793,
241
+ 0.18478454649448395,
242
+ 0.25687918066978443,
243
+ 0.5807053446769717,
244
+ 0.33657354116439797,
245
+ 0.29842707514762895,
246
+ 0.31536528468132013,
247
+ 1.0050891637802128,
248
+ 1.0124161243438718,
249
+ 1.0138051509857173,
250
+ 1.0200439691543575,
251
+ 2.066762685775758,
252
+ 1.4588384628295925,
253
+ 0.5752111673355106,
254
+ 0.25846862792968733,
255
+ 0.34115546941757197,
256
+ 0.7394758462905866,
257
+ 0.4870710670948032,
258
+ 0.3721199631690978,
259
+ 0.5423213839530941,
260
+ 1.1248024702072141,
261
+ 1.1392076015472405,
262
+ 1.1426663398742682,
263
+ 1.301862597465516,
264
+ 2.199875354766848,
265
+ 0.9186902642250119,
266
+ 0.12568794190883642,
267
+ 0.030695030465722098,
268
+ 0.004574332851916548
269
+ ],
270
+ "min": [
271
+ -2.291214942932129,
272
+ -0.001746351015754044,
273
+ -2.4999990463256836,
274
+ -2.814260482788086,
275
+ -2.3611011505126953,
276
+ -1.4937989711761475,
277
+ -1.4969758987426758,
278
+ -1.9597030878067017,
279
+ -1.8635213375091553,
280
+ -1.9709523916244507,
281
+ -1.5,
282
+ -3.0,
283
+ 0.0,
284
+ -2.318650960922241,
285
+ -2.9999561309814453,
286
+ -1.9257696866989136,
287
+ -2.8759055137634277,
288
+ -2.620600461959839,
289
+ -1.5001521110534668,
290
+ -1.4994292259216309,
291
+ -1.5,
292
+ -1.5,
293
+ -1.5,
294
+ -1.5,
295
+ -3.0,
296
+ 4.470348358154297e-08,
297
+ -0.8292319774627686,
298
+ -0.5185094475746155,
299
+ -0.37811079621315
300
+ ],
301
+ "max": [
302
+ 1.4113223552703857,
303
+ 1.8901419639587402,
304
+ 1.7754145860671997,
305
+ 7.309383363462985e-05,
306
+ 2.563594102859497,
307
+ 1.5000243186950684,
308
+ 1.4908421039581299,
309
+ 1.5707963705062866,
310
+ 1.646651268005371,
311
+ 1.709236979484558,
312
+ 2.0620639324188232,
313
+ 3.0,
314
+ 3.0,
315
+ 1.7834906578063965,
316
+ 0.0002448999439366162,
317
+ 2.549729585647583,
318
+ 7.45560391806066e-05,
319
+ 3.0000460147857666,
320
+ 1.4975632429122925,
321
+ 1.4998434782028198,
322
+ 1.5707963705062866,
323
+ 1.7183797359466553,
324
+ 3.979951858520508,
325
+ 3.9879753589630127,
326
+ 3.0,
327
+ 3.0,
328
+ 1.049119234085083,
329
+ 0.6198405623435974,
330
+ 0.45177245140075684
331
+ ],
332
+ "q01": [
333
+ -1.6055770933628097,
334
+ -7.429541994952163e-07,
335
+ -0.9837155520915981,
336
+ -2.480029511451721,
337
+ -0.9457575023174288,
338
+ -0.7192997813224794,
339
+ -1.008323037624359,
340
+ -1.4999999999999987,
341
+ -1.4999999999999987,
342
+ -1.4999999999999987,
343
+ -1.4999999999999987,
344
+ -2.9999999999999973,
345
+ 0.0,
346
+ -1.4375487387180332,
347
+ -1.0682010412216194,
348
+ -0.7535711079835895,
349
+ -2.6383612155914276,
350
+ -0.7471688866615306,
351
+ -0.8604200631380092,
352
+ -1.0900398135185252,
353
+ -1.4999999999999987,
354
+ -1.4999999999999987,
355
+ -1.4999999999999987,
356
+ -1.4999999999999987,
357
+ -2.9999999999999973,
358
+ 0.04336242545396091,
359
+ -0.3402548208832736,
360
+ -0.025130789913237084,
361
+ -0.016220059804618367
362
+ ],
363
+ "q99": [
364
+ 0.6335921049118061,
365
+ 0.8018405169248585,
366
+ 0.3930345878005033,
367
+ -0.06366106234490865,
368
+ 1.0508334636688252,
369
+ 0.8918659299612054,
370
+ 0.5138001590967192,
371
+ 1.4999999999999987,
372
+ 1.4999999999999987,
373
+ 1.4999999999999987,
374
+ 1.4999999999999987,
375
+ 2.9999999999999973,
376
+ 2.9999999999999973,
377
+ 1.1753968834876984,
378
+ -9.999999974752431e-07,
379
+ 0.8976036489009875,
380
+ -0.0579521823674439,
381
+ 1.4876463234424584,
382
+ 0.8774339169263857,
383
+ 1.4641386926174185,
384
+ 1.4999999999999987,
385
+ 1.4999999999999987,
386
+ 1.4999999999999987,
387
+ 1.4999999999999987,
388
+ 2.9999999999999973,
389
+ 2.9999999999999973,
390
+ 0.49598969072103627,
391
+ 0.12137954644858844,
392
+ 0.010726323025301106
393
+ ]
394
+ }
395
+ }
396
+ }
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug_2026_05_26_16_59_09.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c76268dbe3cb4edf581d738f124dbd67c8e00f49a9e8ebc1e5e34a4491e51762
3
+ size 14204515
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/run-metrics.jsonl ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hparams": "{'model': {'type': 'LlavaVLA', 'pretrained_name_or_path': './checkpoints/GR00T-N1.5-3B', 'vlm_backbone': {'type': 'EagleBackbone', 'dtype': 'bf16', 'vlm_path': 'fluxvla/models/third_party_models/eagle2_hg_model', 'vlm_config': {'max_input_seq_len': 900}, 'tune_llm': False, 'tune_visual': True}, 'vla_head': {'type': 'FlowMatchingHead', 'state_dim': 64, 'hidden_size': 1024, 'input_embedding_dim': 1536, 'num_layers': 1, 'num_heads': 4, 'num_inference_timesteps': 4, 'num_steps': 16, 'traj_length': 16, 'action_dim': 32, 'ori_action_dim': 29}, 'freeze_vlm_backbone': False, 'name_mapping': {'vlm_backbone.vlm': 'backbone.eagle_model', 'vla_head': 'action_head'}, 'freeze_projector': False}, '_STAT': 'robocasa_gr1_24tasks_30ep', '_OFFICIAL_GR1_STATS_PATH': 'work_dirs/official_groot_gr1_dataset_statistics.json', 'train_dataloader': {'per_device_batch_size': 16, 'per_device_num_workers': 4, 'dataset': {'type': 'DistributedRepeatingDataset', 'name_mappings': {'observation.state': ['proprio'], 'action': ['action']}, 'statistic_keys': ['observation.state', 'timestamp', 'action'], 'statistic_name': 'robocasa_gr1_24tasks_30ep', 'dataset_statistics_path': 'work_dirs/official_groot_gr1_dataset_statistics.json', 'datasets': [{'type': 'ParquetDataset', 'data_root_path': ['./datasets/robocasa_fluxvla/PnPBottleToCabinetClose', './datasets/robocasa_fluxvla/PnPCanToDrawerClose', './datasets/robocasa_fluxvla/PnPCupToDrawerClose', './datasets/robocasa_fluxvla/PnPMilkToMicrowaveClose', './datasets/robocasa_fluxvla/PnPPotatoToMicrowaveClose', './datasets/robocasa_fluxvla/PnPWineToCabinetClose', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToBasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToPanSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToPotSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToBasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToBowlSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToPlateSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToTieredshelfSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToBowlSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToCardboardboxSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToPanSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToPlateSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToCardboardboxSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToPlateSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToPotSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToTieredbasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToTieredshelfSplitA'], 'statistic_name': 'robocasa_gr1_24tasks_30ep', 'action_key': 'action', 'use_delta': False, 'window_start_idx': 0, 'transforms': [{'type': 'ProcessParquetInputs', 'embodiment_id': 24, 'parquet_keys': ['observation.state', 'timestamp', 'actions', 'info', 'stats', 'action_masks'], 'video_keys': ['observation.images.ego_view'], 'name_mappings': {'observation.state': ['states'], 'actions': ['actions']}}, {'type': 'RobocasaGR1N15Bridge'}, {'type': 'ParquetPrompter'}, {'type': 'ProcessPromptsWithImage', 'max_len': 900, 'num_images': 1, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': 'fluxvla/models/third_party_models/eagle2_hg_model'}}, {'type': 'RandomCropImages', 'scale': 0.95}, {'type': 'ResizeImages', 'height': 224, 'width': 224}, {'type': 'ColorJitterImages', 'brightness': 0.3, 'contrast': 0.4, 'saturation': 0.5, 'hue': 0.08}, {'type': 'NormalizeImages', 'means': [[127.5, 127.5, 127.5]], 'stds': [[127.5, 127.5, 127.5]]}, {'type': 'NormalizeStatesAndActions', 'state_dim': 64, 'action_dim': 32, 'state_key': 'proprio', 'action_key': 'action', 'norm_type': 'min_max', 'normalize_states': False}], 'action_window_size': 16}]}}, 'runner': {'type': 'FSDPTrainRunner', 'max_epochs': None, 'learning_rate': 6e-05, 'weight_decay': 1e-05, 'max_grad_norm': 1.0, 'sampler': None, 'save_iter_interval': 5000, 'save_epoch_interval': 1, 'max_keep_ckpts': 20, 'save_full_model': True, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': 'fluxvla/models/third_party_models/eagle2_hg_model'}, 'collator': {'type': 'DictCollator', 'keys': ['states', 'observation.eepose', 'timestamp', 'images', 'img_masks', 'lang_tokens', 'lang_masks', 'actions', 'action_masks', 'embodiment_ids'], 'meta_keys': ['task_description', 'prompt', 'info', 'stats']}, 'metric': {'type': 'VLAMetric', 'active_trackers': ('jsonl', 'wandb'), 'run_dir': 'work_dirs/groot_robocasa_24x30_0525l_lr6e-5', 'grad_accumulation_steps': 1, 'window_size': 1, 'hparams': Config (path: configs/gr00t/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug.py): {'model': {'type': 'LlavaVLA', 'pretrained_name_or_path': './checkpoints/GR00T-N1.5-3B', 'vlm_backbone': {'type': 'EagleBackbone', 'dtype': 'bf16', 'vlm_path': 'fluxvla/models/third_party_models/eagle2_hg_model', 'vlm_config': {'max_input_seq_len': 900}, 'tune_llm': False, 'tune_visual': True}, 'vla_head': {'type': 'FlowMatchingHead', 'state_dim': 64, 'hidden_size': 1024, 'input_embedding_dim': 1536, 'num_layers': 1, 'num_heads': 4, 'num_inference_timesteps': 4, 'num_steps': 16, 'traj_length': 16, 'action_dim': 32, 'ori_action_dim': 29}, 'freeze_vlm_backbone': False, 'name_mapping': {'vlm_backbone.vlm': 'backbone.eagle_model', 'vla_head': 'action_head'}, 'freeze_projector': False}, '_STAT': 'robocasa_gr1_24tasks_30ep', '_OFFICIAL_GR1_STATS_PATH': 'work_dirs/official_groot_gr1_dataset_statistics.json', 'train_dataloader': {'per_device_batch_size': 16, 'per_device_num_workers': 4, 'dataset': {'type': 'DistributedRepeatingDataset', 'name_mappings': {'observation.state': ['proprio'], 'action': ['action']}, 'statistic_keys': ['observation.state', 'timestamp', 'action'], 'statistic_name': 'robocasa_gr1_24tasks_30ep', 'dataset_statistics_path': 'work_dirs/official_groot_gr1_dataset_statistics.json', 'datasets': [{'type': 'ParquetDataset', 'data_root_path': ['./datasets/robocasa_fluxvla/PnPBottleToCabinetClose', './datasets/robocasa_fluxvla/PnPCanToDrawerClose', './datasets/robocasa_fluxvla/PnPCupToDrawerClose', './datasets/robocasa_fluxvla/PnPMilkToMicrowaveClose', './datasets/robocasa_fluxvla/PnPPotatoToMicrowaveClose', './datasets/robocasa_fluxvla/PnPWineToCabinetClose', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToBasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToPanSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToPotSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToBasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToBowlSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToPlateSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToTieredshelfSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToBowlSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToCardboardboxSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToPanSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToPlateSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToCardboardboxSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToPlateSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToPotSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToTieredbasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToTieredshelfSplitA'], 'statistic_name': 'robocasa_gr1_24tasks_30ep', 'action_key': 'action', 'use_delta': False, 'window_start_idx': 0, 'transforms': [{'type': 'ProcessParquetInputs', 'embodiment_id': 24, 'parquet_keys': ['observation.state', 'timestamp', 'actions', 'info', 'stats', 'action_masks'], 'video_keys': ['observation.images.ego_view'], 'name_mappings': {'observation.state': ['states'], 'actions': ['actions']}}, {'type': 'RobocasaGR1N15Bridge'}, {'type': 'ParquetPrompter'}, {'type': 'ProcessPromptsWithImage', 'max_len': 900, 'num_images': 1, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': 'fluxvla/models/third_party_models/eagle2_hg_model'}}, {'type': 'RandomCropImages', 'scale': 0.95}, {'type': 'ResizeImages', 'height': 224, 'width': 224}, {'type': 'ColorJitterImages', 'brightness': 0.3, 'contrast': 0.4, 'saturation': 0.5, 'hue': 0.08}, {'type': 'NormalizeImages', 'means': [[127.5, 127.5, 127.5]], 'stds': [[127.5, 127.5, 127.5]]}, {'type': 'NormalizeStatesAndActions', 'state_dim': 64, 'action_dim': 32, 'state_key': 'proprio', 'action_key': 'action', 'norm_type': 'min_max', 'normalize_states': False}], 'action_window_size': 16}]}}, 'runner': {'type': 'FSDPTrainRunner', 'max_epochs': None, 'learning_rate': 6e-05, 'weight_decay': 1e-05, 'max_grad_norm': 1.0, 'sampler': None, 'save_iter_interval': 5000, 'save_epoch_interval': 1, 'max_keep_ckpts': 20, 'save_full_model': True, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': 'fluxvla/models/third_party_models/eagle2_hg_model'}, 'collator': {'type': 'DictCollator', 'keys': ['states', 'observation.eepose', 'timestamp', 'images', 'img_masks', 'lang_tokens', 'lang_masks', 'actions', 'action_masks', 'embodiment_ids'], 'meta_keys': ['task_description', 'prompt', 'info', 'stats']}, 'metric': {'type': 'VLAMetric', 'active_trackers': ('jsonl', 'wandb'), 'run_dir': 'work_dirs/groot_robocasa_24x30_0525l_lr6e-5', 'grad_accumulation_steps': 1, 'window_size': 1, 'hparams': Config (path: configs/gr00t/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug.py): {...}, 'run_id': 'gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug_2026_05_26_16_59_09'}, 'lr_scheduler_type': 'linear-warmup+cosine-decay', 'warmup_ratio': 0.05, 'enable_gradient_checkpointing': False, 'enable_mixed_precision_training': True, 'mixed_precision_dtype': 'bf16', 'sharding_strategy': 'no-shard', 'change_key_name': False, 'max_steps': 50000, 'cfg': Config (path: configs/gr00t/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug.py): {...}, 'args': Namespace(config='configs/gr00t/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug.py', work_dir='work_dirs/groot_robocasa_24x30_0525l_lr6e-5', cfg_options={'runner.type': 'FSDPTrainRunner', 'runner.sharding_strategy': 'no-shard', 'train_dataloader.per_device_batch_size': 16, 'runner.enable_gradient_checkpointing': False, 'runner.learning_rate': 6e-05, 'runner.weight_decay': 1e-05, 'runner.warmup_ratio': 0.05, 'runner.lr_scheduler_type': 'linear-warmup+cosine-decay', 'runner.save_iter_interval': 5000, 'runner.save_epoch_interval': 1, 'runner.max_keep_ckpts': 20, 'runner.metric.active_trackers': ('jsonl', 'wandb'), 'runner.max_epochs': None, 'runner.max_steps': 50000}, eval_after_train=False, resume_from=None)}, 'inference_model': {'type': 'LlavaVLA', 'pretrained_name_or_path': './checkpoints/GR00T-N1.5-3B', 'vlm_backbone': {'type': 'EagleInferenceBackbone', 'vlm_path': 'fluxvla/models/third_party_models/eagle2_hg_model', 'vlm_config': {'max_input_seq_len': 900}}, 'vla_head': {'type': 'FlowMatchingInferenceHead', 'state_dim': 64, 'hidden_size': 1024, 'input_embedding_dim': 1536, 'num_layers': 1, 'num_heads': 4, 'num_steps': 16, 'num_inference_timesteps': 4, 'traj_length': 16, 'ori_action_dim': 29, 'action_dim': 32, 'max_input_seq_len': 900, 'diffusion_model_cfg': {'attention_head_dim': 48, 'cross_attention_dim': 2048, 'dropout': 0.2, 'final_dropout': True, 'interleave_self_attention': True, 'norm_type': 'ada_norm', 'num_attention_heads': 32, 'num_layers': 16, 'output_dim': 1024, 'positional_embeddings': None}}}, 'eval': {'type': 'RobocasaEvalRunner', 'model_family': 'groot', 'task_list': ['gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env'], 'eval_chunk_size': 16, 'max_episode_steps': 720, 'num_trials_per_task': 20, 'seed': 7, 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'dataset': {'type': 'RobocasaEvalDataset', 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'transforms': [{'type': 'ProcessRobocasaEvalInputs', 'img_key': 'video.ego_view_bg_crop_pad_res256_freq20', 'resize_size': 224, 'center_crop_scale': 0.95, 'normalize': False, 'embodiment_id': 24}, {'type': 'RobocasaGR1N15Bridge'}, {'type': 'NormalizeStatesAndActions', 'state_dim': 64, 'state_key': 'proprio', 'action_key': 'action', 'norm_type': 'min_max', 'normalize_states': False}, {'type': 'ProcessPromptsWithImage', 'max_len': 900, 'num_images': 1, 'return_text': True, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': 'fluxvla/models/third_party_models/eagle2_hg_model'}}, {'type': 'NormalizeImages', 'means': [[127.5, 127.5, 127.5]], 'stds': [[127.5, 127.5, 127.5]]}]}, 'denormalize_action': {'type': 'DenormalizeRobocasaAction', 'norm_type': 'min_max', 'action_dim': 29, 'clip_actions': False, 'stats_order': 'fluxvla'}}}, 'run_id': 'gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug_2026_05_26_16_59_09'}, 'lr_scheduler_type': 'linear-warmup+cosine-decay', 'warmup_ratio': 0.05, 'enable_gradient_checkpointing': False, 'enable_mixed_precision_training': True, 'mixed_precision_dtype': 'bf16', 'sharding_strategy': 'no-shard', 'change_key_name': False, 'max_steps': 50000, 'cfg': Config (path: configs/gr00t/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug.py): {'model': {'type': 'LlavaVLA', 'pretrained_name_or_path': './checkpoints/GR00T-N1.5-3B', 'vlm_backbone': {'type': 'EagleBackbone', 'dtype': 'bf16', 'vlm_path': 'fluxvla/models/third_party_models/eagle2_hg_model', 'vlm_config': {'max_input_seq_len': 900}, 'tune_llm': False, 'tune_visual': True}, 'vla_head': {'type': 'FlowMatchingHead', 'state_dim': 64, 'hidden_size': 1024, 'input_embedding_dim': 1536, 'num_layers': 1, 'num_heads': 4, 'num_inference_timesteps': 4, 'num_steps': 16, 'traj_length': 16, 'action_dim': 32, 'ori_action_dim': 29}, 'freeze_vlm_backbone': False, 'name_mapping': {'vlm_backbone.vlm': 'backbone.eagle_model', 'vla_head': 'action_head'}, 'freeze_projector': False}, '_STAT': 'robocasa_gr1_24tasks_30ep', '_OFFICIAL_GR1_STATS_PATH': 'work_dirs/official_groot_gr1_dataset_statistics.json', 'train_dataloader': {'per_device_batch_size': 16, 'per_device_num_workers': 4, 'dataset': {'type': 'DistributedRepeatingDataset', 'name_mappings': {'observation.state': ['proprio'], 'action': ['action']}, 'statistic_keys': ['observation.state', 'timestamp', 'action'], 'statistic_name': 'robocasa_gr1_24tasks_30ep', 'dataset_statistics_path': 'work_dirs/official_groot_gr1_dataset_statistics.json', 'datasets': [{'type': 'ParquetDataset', 'data_root_path': ['./datasets/robocasa_fluxvla/PnPBottleToCabinetClose', './datasets/robocasa_fluxvla/PnPCanToDrawerClose', './datasets/robocasa_fluxvla/PnPCupToDrawerClose', './datasets/robocasa_fluxvla/PnPMilkToMicrowaveClose', './datasets/robocasa_fluxvla/PnPPotatoToMicrowaveClose', './datasets/robocasa_fluxvla/PnPWineToCabinetClose', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToBasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToPanSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToPotSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToBasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToBowlSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToPlateSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlacematToTieredshelfSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToBowlSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToCardboardboxSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToPanSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromPlateToPlateSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToCardboardboxSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToPlateSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToPotSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToTieredbasketSplitA', './datasets/robocasa_fluxvla/PosttrainPnPNovelFromTrayToTieredshelfSplitA'], 'statistic_name': 'robocasa_gr1_24tasks_30ep', 'action_key': 'action', 'use_delta': False, 'window_start_idx': 0, 'transforms': [{'type': 'ProcessParquetInputs', 'embodiment_id': 24, 'parquet_keys': ['observation.state', 'timestamp', 'actions', 'info', 'stats', 'action_masks'], 'video_keys': ['observation.images.ego_view'], 'name_mappings': {'observation.state': ['states'], 'actions': ['actions']}}, {'type': 'RobocasaGR1N15Bridge'}, {'type': 'ParquetPrompter'}, {'type': 'ProcessPromptsWithImage', 'max_len': 900, 'num_images': 1, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': 'fluxvla/models/third_party_models/eagle2_hg_model'}}, {'type': 'RandomCropImages', 'scale': 0.95}, {'type': 'ResizeImages', 'height': 224, 'width': 224}, {'type': 'ColorJitterImages', 'brightness': 0.3, 'contrast': 0.4, 'saturation': 0.5, 'hue': 0.08}, {'type': 'NormalizeImages', 'means': [[127.5, 127.5, 127.5]], 'stds': [[127.5, 127.5, 127.5]]}, {'type': 'NormalizeStatesAndActions', 'state_dim': 64, 'action_dim': 32, 'state_key': 'proprio', 'action_key': 'action', 'norm_type': 'min_max', 'normalize_states': False}], 'action_window_size': 16}]}}, 'runner': {'type': 'FSDPTrainRunner', 'max_epochs': None, 'learning_rate': 6e-05, 'weight_decay': 1e-05, 'max_grad_norm': 1.0, 'sampler': None, 'save_iter_interval': 5000, 'save_epoch_interval': 1, 'max_keep_ckpts': 20, 'save_full_model': True, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': 'fluxvla/models/third_party_models/eagle2_hg_model'}, 'collator': {'type': 'DictCollator', 'keys': ['states', 'observation.eepose', 'timestamp', 'images', 'img_masks', 'lang_tokens', 'lang_masks', 'actions', 'action_masks', 'embodiment_ids'], 'meta_keys': ['task_description', 'prompt', 'info', 'stats']}, 'metric': {'type': 'VLAMetric', 'active_trackers': ('jsonl', 'wandb'), 'run_dir': 'work_dirs/groot_robocasa_24x30_0525l_lr6e-5', 'grad_accumulation_steps': 1, 'window_size': 1, 'hparams': Config (path: configs/gr00t/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug.py): {...}, 'run_id': 'gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug_2026_05_26_16_59_09'}, 'lr_scheduler_type': 'linear-warmup+cosine-decay', 'warmup_ratio': 0.05, 'enable_gradient_checkpointing': False, 'enable_mixed_precision_training': True, 'mixed_precision_dtype': 'bf16', 'sharding_strategy': 'no-shard', 'change_key_name': False, 'max_steps': 50000, 'cfg': Config (path: configs/gr00t/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug.py): {...}, 'args': Namespace(config='configs/gr00t/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug.py', work_dir='work_dirs/groot_robocasa_24x30_0525l_lr6e-5', cfg_options={'runner.type': 'FSDPTrainRunner', 'runner.sharding_strategy': 'no-shard', 'train_dataloader.per_device_batch_size': 16, 'runner.enable_gradient_checkpointing': False, 'runner.learning_rate': 6e-05, 'runner.weight_decay': 1e-05, 'runner.warmup_ratio': 0.05, 'runner.lr_scheduler_type': 'linear-warmup+cosine-decay', 'runner.save_iter_interval': 5000, 'runner.save_epoch_interval': 1, 'runner.max_keep_ckpts': 20, 'runner.metric.active_trackers': ('jsonl', 'wandb'), 'runner.max_epochs': None, 'runner.max_steps': 50000}, eval_after_train=False, resume_from=None)}, 'inference_model': {'type': 'LlavaVLA', 'pretrained_name_or_path': './checkpoints/GR00T-N1.5-3B', 'vlm_backbone': {'type': 'EagleInferenceBackbone', 'vlm_path': 'fluxvla/models/third_party_models/eagle2_hg_model', 'vlm_config': {'max_input_seq_len': 900}}, 'vla_head': {'type': 'FlowMatchingInferenceHead', 'state_dim': 64, 'hidden_size': 1024, 'input_embedding_dim': 1536, 'num_layers': 1, 'num_heads': 4, 'num_steps': 16, 'num_inference_timesteps': 4, 'traj_length': 16, 'ori_action_dim': 29, 'action_dim': 32, 'max_input_seq_len': 900, 'diffusion_model_cfg': {'attention_head_dim': 48, 'cross_attention_dim': 2048, 'dropout': 0.2, 'final_dropout': True, 'interleave_self_attention': True, 'norm_type': 'ada_norm', 'num_attention_heads': 32, 'num_layers': 16, 'output_dim': 1024, 'positional_embeddings': None}}}, 'eval': {'type': 'RobocasaEvalRunner', 'model_family': 'groot', 'task_list': ['gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env'], 'eval_chunk_size': 16, 'max_episode_steps': 720, 'num_trials_per_task': 20, 'seed': 7, 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'dataset': {'type': 'RobocasaEvalDataset', 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'transforms': [{'type': 'ProcessRobocasaEvalInputs', 'img_key': 'video.ego_view_bg_crop_pad_res256_freq20', 'resize_size': 224, 'center_crop_scale': 0.95, 'normalize': False, 'embodiment_id': 24}, {'type': 'RobocasaGR1N15Bridge'}, {'type': 'NormalizeStatesAndActions', 'state_dim': 64, 'state_key': 'proprio', 'action_key': 'action', 'norm_type': 'min_max', 'normalize_states': False}, {'type': 'ProcessPromptsWithImage', 'max_len': 900, 'num_images': 1, 'return_text': True, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': 'fluxvla/models/third_party_models/eagle2_hg_model'}}, {'type': 'NormalizeImages', 'means': [[127.5, 127.5, 127.5]], 'stds': [[127.5, 127.5, 127.5]]}]}, 'denormalize_action': {'type': 'DenormalizeRobocasaAction', 'norm_type': 'min_max', 'action_dim': 29, 'clip_actions': False, 'stats_order': 'fluxvla'}}}, 'args': Namespace(config='configs/gr00t/gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug.py', work_dir='work_dirs/groot_robocasa_24x30_0525l_lr6e-5', cfg_options={'runner.type': 'FSDPTrainRunner', 'runner.sharding_strategy': 'no-shard', 'train_dataloader.per_device_batch_size': 16, 'runner.enable_gradient_checkpointing': False, 'runner.learning_rate': 6e-05, 'runner.weight_decay': 1e-05, 'runner.warmup_ratio': 0.05, 'runner.lr_scheduler_type': 'linear-warmup+cosine-decay', 'runner.save_iter_interval': 5000, 'runner.save_epoch_interval': 1, 'runner.max_keep_ckpts': 20, 'runner.metric.active_trackers': ('jsonl', 'wandb'), 'runner.max_epochs': None, 'runner.max_steps': 50000}, eval_after_train=False, resume_from=None)}, 'inference_model': {'type': 'LlavaVLA', 'pretrained_name_or_path': './checkpoints/GR00T-N1.5-3B', 'vlm_backbone': {'type': 'EagleInferenceBackbone', 'vlm_path': 'fluxvla/models/third_party_models/eagle2_hg_model', 'vlm_config': {'max_input_seq_len': 900}}, 'vla_head': {'type': 'FlowMatchingInferenceHead', 'state_dim': 64, 'hidden_size': 1024, 'input_embedding_dim': 1536, 'num_layers': 1, 'num_heads': 4, 'num_steps': 16, 'num_inference_timesteps': 4, 'traj_length': 16, 'ori_action_dim': 29, 'action_dim': 32, 'max_input_seq_len': 900, 'diffusion_model_cfg': {'attention_head_dim': 48, 'cross_attention_dim': 2048, 'dropout': 0.2, 'final_dropout': True, 'interleave_self_attention': True, 'norm_type': 'ada_norm', 'num_attention_heads': 32, 'num_layers': 16, 'output_dim': 1024, 'positional_embeddings': None}}}, 'eval': {'type': 'RobocasaEvalRunner', 'model_family': 'groot', 'task_list': ['gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env'], 'eval_chunk_size': 16, 'max_episode_steps': 720, 'num_trials_per_task': 20, 'seed': 7, 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'dataset': {'type': 'RobocasaEvalDataset', 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'transforms': [{'type': 'ProcessRobocasaEvalInputs', 'img_key': 'video.ego_view_bg_crop_pad_res256_freq20', 'resize_size': 224, 'center_crop_scale': 0.95, 'normalize': False, 'embodiment_id': 24}, {'type': 'RobocasaGR1N15Bridge'}, {'type': 'NormalizeStatesAndActions', 'state_dim': 64, 'state_key': 'proprio', 'action_key': 'action', 'norm_type': 'min_max', 'normalize_states': False}, {'type': 'ProcessPromptsWithImage', 'max_len': 900, 'num_images': 1, 'return_text': True, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': 'fluxvla/models/third_party_models/eagle2_hg_model'}}, {'type': 'NormalizeImages', 'means': [[127.5, 127.5, 127.5]], 'stds': [[127.5, 127.5, 127.5]]}]}, 'denormalize_action': {'type': 'DenormalizeRobocasaAction', 'norm_type': 'min_max', 'action_dim': 29, 'clip_actions': False, 'stats_order': 'fluxvla'}}}", "run_id": "gr00t_eagle_3b_robocasa_posttrain_24x30_official_aug_2026_05_26_16_59_09"}
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/tokenizer/added_tokens.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</box>": 151673,
3
+ "</img>": 151671,
4
+ "</interval>": 151679,
5
+ "</quad>": 151675,
6
+ "</ref>": 151677,
7
+ "</think>": 151668,
8
+ "</tool_call>": 151658,
9
+ "</tool_response>": 151666,
10
+ "<IMG_CONTEXT>": 151669,
11
+ "<box>": 151672,
12
+ "<img>": 151670,
13
+ "<interval>": 151678,
14
+ "<quad>": 151674,
15
+ "<ref>": 151676,
16
+ "<think>": 151667,
17
+ "<tool_call>": 151657,
18
+ "<tool_response>": 151665,
19
+ "<|box_end|>": 151649,
20
+ "<|box_start|>": 151648,
21
+ "<|endoftext|>": 151643,
22
+ "<|file_sep|>": 151664,
23
+ "<|fim_middle|>": 151660,
24
+ "<|fim_pad|>": 151662,
25
+ "<|fim_prefix|>": 151659,
26
+ "<|fim_suffix|>": 151661,
27
+ "<|im_end|>": 151645,
28
+ "<|im_start|>": 151644,
29
+ "<|image_pad|>": 151655,
30
+ "<|object_ref_end|>": 151647,
31
+ "<|object_ref_start|>": 151646,
32
+ "<|quad_end|>": 151651,
33
+ "<|quad_start|>": 151650,
34
+ "<|repo_name|>": 151663,
35
+ "<|video_pad|>": 151656,
36
+ "<|vision_end|>": 151653,
37
+ "<|vision_pad|>": 151654,
38
+ "<|vision_start|>": 151652
39
+ }
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/tokenizer/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/tokenizer/special_tokens_map.json ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|object_ref_start|>",
6
+ "<|object_ref_end|>",
7
+ "<|box_start|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_start|>",
12
+ "<|vision_end|>",
13
+ "<|vision_pad|>",
14
+ "<|image_pad|>",
15
+ "<|video_pad|>",
16
+ "<IMG_CONTEXT>",
17
+ "<img>",
18
+ "</img>",
19
+ "<box>",
20
+ "</box>",
21
+ "<quad>",
22
+ "</quad>",
23
+ "<ref>",
24
+ "</ref>",
25
+ "<interval>",
26
+ "</interval>"
27
+ ],
28
+ "eos_token": {
29
+ "content": "<|im_end|>",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false
34
+ },
35
+ "pad_token": {
36
+ "content": "<|endoftext|>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false
41
+ }
42
+ }
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,344 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_eos_token": false,
4
+ "add_prefix_space": false,
5
+ "added_tokens_decoder": {
6
+ "151643": {
7
+ "content": "<|endoftext|>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false,
12
+ "special": true
13
+ },
14
+ "151644": {
15
+ "content": "<|im_start|>",
16
+ "lstrip": false,
17
+ "normalized": false,
18
+ "rstrip": false,
19
+ "single_word": false,
20
+ "special": true
21
+ },
22
+ "151645": {
23
+ "content": "<|im_end|>",
24
+ "lstrip": false,
25
+ "normalized": false,
26
+ "rstrip": false,
27
+ "single_word": false,
28
+ "special": true
29
+ },
30
+ "151646": {
31
+ "content": "<|object_ref_start|>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false,
36
+ "special": true
37
+ },
38
+ "151647": {
39
+ "content": "<|object_ref_end|>",
40
+ "lstrip": false,
41
+ "normalized": false,
42
+ "rstrip": false,
43
+ "single_word": false,
44
+ "special": true
45
+ },
46
+ "151648": {
47
+ "content": "<|box_start|>",
48
+ "lstrip": false,
49
+ "normalized": false,
50
+ "rstrip": false,
51
+ "single_word": false,
52
+ "special": true
53
+ },
54
+ "151649": {
55
+ "content": "<|box_end|>",
56
+ "lstrip": false,
57
+ "normalized": false,
58
+ "rstrip": false,
59
+ "single_word": false,
60
+ "special": true
61
+ },
62
+ "151650": {
63
+ "content": "<|quad_start|>",
64
+ "lstrip": false,
65
+ "normalized": false,
66
+ "rstrip": false,
67
+ "single_word": false,
68
+ "special": true
69
+ },
70
+ "151651": {
71
+ "content": "<|quad_end|>",
72
+ "lstrip": false,
73
+ "normalized": false,
74
+ "rstrip": false,
75
+ "single_word": false,
76
+ "special": true
77
+ },
78
+ "151652": {
79
+ "content": "<|vision_start|>",
80
+ "lstrip": false,
81
+ "normalized": false,
82
+ "rstrip": false,
83
+ "single_word": false,
84
+ "special": true
85
+ },
86
+ "151653": {
87
+ "content": "<|vision_end|>",
88
+ "lstrip": false,
89
+ "normalized": false,
90
+ "rstrip": false,
91
+ "single_word": false,
92
+ "special": true
93
+ },
94
+ "151654": {
95
+ "content": "<|vision_pad|>",
96
+ "lstrip": false,
97
+ "normalized": false,
98
+ "rstrip": false,
99
+ "single_word": false,
100
+ "special": true
101
+ },
102
+ "151655": {
103
+ "content": "<|image_pad|>",
104
+ "lstrip": false,
105
+ "normalized": false,
106
+ "rstrip": false,
107
+ "single_word": false,
108
+ "special": true
109
+ },
110
+ "151656": {
111
+ "content": "<|video_pad|>",
112
+ "lstrip": false,
113
+ "normalized": false,
114
+ "rstrip": false,
115
+ "single_word": false,
116
+ "special": true
117
+ },
118
+ "151657": {
119
+ "content": "<tool_call>",
120
+ "lstrip": false,
121
+ "normalized": false,
122
+ "rstrip": false,
123
+ "single_word": false,
124
+ "special": false
125
+ },
126
+ "151658": {
127
+ "content": "</tool_call>",
128
+ "lstrip": false,
129
+ "normalized": false,
130
+ "rstrip": false,
131
+ "single_word": false,
132
+ "special": false
133
+ },
134
+ "151659": {
135
+ "content": "<|fim_prefix|>",
136
+ "lstrip": false,
137
+ "normalized": false,
138
+ "rstrip": false,
139
+ "single_word": false,
140
+ "special": false
141
+ },
142
+ "151660": {
143
+ "content": "<|fim_middle|>",
144
+ "lstrip": false,
145
+ "normalized": false,
146
+ "rstrip": false,
147
+ "single_word": false,
148
+ "special": false
149
+ },
150
+ "151661": {
151
+ "content": "<|fim_suffix|>",
152
+ "lstrip": false,
153
+ "normalized": false,
154
+ "rstrip": false,
155
+ "single_word": false,
156
+ "special": false
157
+ },
158
+ "151662": {
159
+ "content": "<|fim_pad|>",
160
+ "lstrip": false,
161
+ "normalized": false,
162
+ "rstrip": false,
163
+ "single_word": false,
164
+ "special": false
165
+ },
166
+ "151663": {
167
+ "content": "<|repo_name|>",
168
+ "lstrip": false,
169
+ "normalized": false,
170
+ "rstrip": false,
171
+ "single_word": false,
172
+ "special": false
173
+ },
174
+ "151664": {
175
+ "content": "<|file_sep|>",
176
+ "lstrip": false,
177
+ "normalized": false,
178
+ "rstrip": false,
179
+ "single_word": false,
180
+ "special": false
181
+ },
182
+ "151665": {
183
+ "content": "<tool_response>",
184
+ "lstrip": false,
185
+ "normalized": false,
186
+ "rstrip": false,
187
+ "single_word": false,
188
+ "special": false
189
+ },
190
+ "151666": {
191
+ "content": "</tool_response>",
192
+ "lstrip": false,
193
+ "normalized": false,
194
+ "rstrip": false,
195
+ "single_word": false,
196
+ "special": false
197
+ },
198
+ "151667": {
199
+ "content": "<think>",
200
+ "lstrip": false,
201
+ "normalized": false,
202
+ "rstrip": false,
203
+ "single_word": false,
204
+ "special": false
205
+ },
206
+ "151668": {
207
+ "content": "</think>",
208
+ "lstrip": false,
209
+ "normalized": false,
210
+ "rstrip": false,
211
+ "single_word": false,
212
+ "special": false
213
+ },
214
+ "151669": {
215
+ "content": "<IMG_CONTEXT>",
216
+ "lstrip": false,
217
+ "normalized": false,
218
+ "rstrip": false,
219
+ "single_word": false,
220
+ "special": true
221
+ },
222
+ "151670": {
223
+ "content": "<img>",
224
+ "lstrip": false,
225
+ "normalized": false,
226
+ "rstrip": false,
227
+ "single_word": false,
228
+ "special": true
229
+ },
230
+ "151671": {
231
+ "content": "</img>",
232
+ "lstrip": false,
233
+ "normalized": false,
234
+ "rstrip": false,
235
+ "single_word": false,
236
+ "special": true
237
+ },
238
+ "151672": {
239
+ "content": "<box>",
240
+ "lstrip": false,
241
+ "normalized": false,
242
+ "rstrip": false,
243
+ "single_word": false,
244
+ "special": true
245
+ },
246
+ "151673": {
247
+ "content": "</box>",
248
+ "lstrip": false,
249
+ "normalized": false,
250
+ "rstrip": false,
251
+ "single_word": false,
252
+ "special": true
253
+ },
254
+ "151674": {
255
+ "content": "<quad>",
256
+ "lstrip": false,
257
+ "normalized": false,
258
+ "rstrip": false,
259
+ "single_word": false,
260
+ "special": true
261
+ },
262
+ "151675": {
263
+ "content": "</quad>",
264
+ "lstrip": false,
265
+ "normalized": false,
266
+ "rstrip": false,
267
+ "single_word": false,
268
+ "special": true
269
+ },
270
+ "151676": {
271
+ "content": "<ref>",
272
+ "lstrip": false,
273
+ "normalized": false,
274
+ "rstrip": false,
275
+ "single_word": false,
276
+ "special": true
277
+ },
278
+ "151677": {
279
+ "content": "</ref>",
280
+ "lstrip": false,
281
+ "normalized": false,
282
+ "rstrip": false,
283
+ "single_word": false,
284
+ "special": true
285
+ },
286
+ "151678": {
287
+ "content": "<interval>",
288
+ "lstrip": false,
289
+ "normalized": false,
290
+ "rstrip": false,
291
+ "single_word": false,
292
+ "special": true
293
+ },
294
+ "151679": {
295
+ "content": "</interval>",
296
+ "lstrip": false,
297
+ "normalized": false,
298
+ "rstrip": false,
299
+ "single_word": false,
300
+ "special": true
301
+ }
302
+ },
303
+ "additional_special_tokens": [
304
+ "<|im_start|>",
305
+ "<|im_end|>",
306
+ "<|object_ref_start|>",
307
+ "<|object_ref_end|>",
308
+ "<|box_start|>",
309
+ "<|box_end|>",
310
+ "<|quad_start|>",
311
+ "<|quad_end|>",
312
+ "<|vision_start|>",
313
+ "<|vision_end|>",
314
+ "<|vision_pad|>",
315
+ "<|image_pad|>",
316
+ "<|video_pad|>",
317
+ "<IMG_CONTEXT>",
318
+ "<img>",
319
+ "</img>",
320
+ "<box>",
321
+ "</box>",
322
+ "<quad>",
323
+ "</quad>",
324
+ "<ref>",
325
+ "</ref>",
326
+ "<interval>",
327
+ "</interval>"
328
+ ],
329
+ "auto_map": {
330
+ "AutoProcessor": "processing_eagle2_5_vl.Eagle2_5_VLProcessor"
331
+ },
332
+ "bos_token": null,
333
+ "chat_template": "{%- if tools %}\n {{- '<|im_start|>system\\n' }}\n {%- if messages[0].role == 'system' %}\n {{- messages[0].content + '\\n\\n' }}\n {%- endif %}\n {{- \"# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n {%- for tool in tools %}\n {{- \"\\n\" }}\n {{- tool | tojson }}\n {%- endfor %}\n {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n {%- if messages[0].role == 'system' %}\n {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n {%- set index = (messages|length - 1) - loop.index0 %}\n {%- if ns.multi_step_tool and message.role == \"user\" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n {%- set ns.multi_step_tool = false %}\n {%- set ns.last_query_index = index %}\n {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {%- set content = message.content %}\n {%- set reasoning_content = '' %}\n {%- if message.reasoning_content is defined and message.reasoning_content is not none %}\n {%- set reasoning_content = message.reasoning_content %}\n {%- else %}\n {%- if '</think>' in message.content %}\n {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n {%- endif %}\n {%- endif %}\n {%- if loop.index0 > ns.last_query_index %}\n {%- if loop.last or (not loop.last and reasoning_content) %}\n {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- if message.tool_calls %}\n {%- for tool_call in message.tool_calls %}\n {%- if (loop.first and content) or (not loop.first) %}\n {{- '\\n' }}\n {%- endif %}\n {%- if tool_call.function %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '<tool_call>\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {%- if tool_call.arguments is string %}\n {{- tool_call.arguments }}\n {%- else %}\n {{- tool_call.arguments | tojson }}\n {%- endif %}\n {{- '}\\n</tool_call>' }}\n {%- endfor %}\n {%- endif %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- message.content }}\n {{- '\\n</tool_response>' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n {%- if enable_thinking is defined and enable_thinking is false %}\n {{- '<think>\\n\\n</think>\\n\\n' }}\n {%- endif %}\n{%- endif %}",
334
+ "clean_up_tokenization_spaces": false,
335
+ "eos_token": "<|im_end|>",
336
+ "errors": "replace",
337
+ "extra_special_tokens": {},
338
+ "model_max_length": 16384,
339
+ "pad_token": "<|endoftext|>",
340
+ "processor_class": "Eagle2_5_VLProcessor",
341
+ "split_special_tokens": false,
342
+ "tokenizer_class": "Qwen2Tokenizer",
343
+ "unk_token": null
344
+ }
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/tokenizer/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
gr00t_eagle_3b_robocasa_gr1_24x30_finetune_bs64/vlm_backbone_config.json ADDED
@@ -0,0 +1,106 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_attn_implementation": "flash_attention_2",
3
+ "architectures": [
4
+ "Eagle2_5_VLForConditionalGeneration"
5
+ ],
6
+ "auto_map": {
7
+ "AutoConfig": "configuration_eagle2_5_vl.Eagle2_5_VLConfig",
8
+ "AutoModel": "modeling_eagle2_5_vl.Eagle2_5_VLForConditionalGeneration"
9
+ },
10
+ "downsample_ratio": 0.5,
11
+ "dynamic_image_size": true,
12
+ "force_image_size": 224,
13
+ "image_token_index": 151669,
14
+ "initializer_range": 0.02,
15
+ "loss_version": "efficient_v2_cp_head",
16
+ "max_dynamic_tiles": 12,
17
+ "min_dynamic_tiles": 1,
18
+ "mlp_checkpoint": false,
19
+ "mlp_connector_layers": 1,
20
+ "model_type": "eagle_2_5_vl",
21
+ "output_attentions": false,
22
+ "pad2square": false,
23
+ "select_layer": -1,
24
+ "template": "qwen3-chat",
25
+ "text_config": {
26
+ "_name_or_path": "Qwen/Qwen3-1.7B",
27
+ "architectures": [
28
+ "Qwen3ForCausalLM"
29
+ ],
30
+ "attention_bias": false,
31
+ "attention_dropout": 0,
32
+ "bos_token_id": 151643,
33
+ "eos_token_id": 151645,
34
+ "head_dim": 128,
35
+ "hidden_act": "silu",
36
+ "hidden_size": 2048,
37
+ "initializer_range": 0.02,
38
+ "intermediate_size": 6144,
39
+ "layer_types": [
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention",
52
+ "full_attention",
53
+ "full_attention",
54
+ "full_attention",
55
+ "full_attention",
56
+ "full_attention",
57
+ "full_attention",
58
+ "full_attention",
59
+ "full_attention",
60
+ "full_attention",
61
+ "full_attention",
62
+ "full_attention",
63
+ "full_attention",
64
+ "full_attention",
65
+ "full_attention",
66
+ "full_attention",
67
+ "full_attention"
68
+ ],
69
+ "max_position_embeddings": 40960,
70
+ "max_window_layers": 28,
71
+ "model_type": "qwen3",
72
+ "num_attention_heads": 16,
73
+ "num_hidden_layers": 12,
74
+ "num_key_value_heads": 8,
75
+ "rms_norm_eps": 1e-06,
76
+ "rope_scaling": null,
77
+ "rope_theta": 1000000,
78
+ "sliding_window": null,
79
+ "tie_word_embeddings": true,
80
+ "torch_dtype": "bfloat16",
81
+ "use_cache": false,
82
+ "use_sliding_window": false,
83
+ "vocab_size": 151680
84
+ },
85
+ "tie_word_embeddings": true,
86
+ "torch_dtype": "bfloat16",
87
+ "transformers_version": null,
88
+ "use_backbone_lora": 0,
89
+ "use_llm_lora": 0,
90
+ "use_pixel_shuffle": false,
91
+ "use_thumbnail": true,
92
+ "vision_config": {
93
+ "attention_dropout": 0,
94
+ "hidden_act": "gelu_pytorch_tanh",
95
+ "hidden_size": 1152,
96
+ "image_size": 224,
97
+ "intermediate_size": 4304,
98
+ "layer_norm_eps": 1e-06,
99
+ "model_type": "siglip_vision_model",
100
+ "num_attention_heads": 16,
101
+ "num_channels": 3,
102
+ "num_hidden_layers": 27,
103
+ "patch_size": 14,
104
+ "torch_dtype": "bfloat16"
105
+ }
106
+ }