Hugging Face's logo

RobotisSW
/
Task_802_groot_100k

Robotics
Safetensors
Gr00tN1d7
robotis
cyclo_intelligence
Model card Files
xet
 Community
RobotisSW commited on
Commit
cf7a20a
·
verified ·
1 Parent(s): e482e91

Upload folder using huggingface_hub

Browse files
Files changed (20) hide show
  1. README.md +11 -0
  2. config.json +96 -0
  3. embodiment_id.json +57 -0
  4. experiment_cfg/conf.yaml +232 -0
  5. experiment_cfg/config.yaml +230 -0
  6. experiment_cfg/dataset_statistics.json +769 -0
  7. experiment_cfg/final_model_config.json +64 -0
  8. experiment_cfg/final_processor_config.json +0 -0
  9. model-00001-of-00003.safetensors +3 -0
  10. model-00002-of-00003.safetensors +3 -0
  11. model-00003-of-00003.safetensors +3 -0
  12. model.safetensors.index.json +0 -0
  13. optimizer.pt +3 -0
  14. processor_config.json +1129 -0
  15. rng_state.pth +3 -0
  16. scheduler.pt +3 -0
  17. statistics.json +0 -0
  18. trainer_state.json +0 -0
  19. training_args.bin +3 -0
  20. wandb_config.json +1 -0
README.md ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ pipeline_tag: robotics
3
+ tags:
4
+ - robotis
5
+ - cyclo_intelligence
6
+ - robotics
7
+ ---
8
+
9
+ # checkpoint-100000
10
+
11
+ Created with [Cyclo Intelligence](https://github.com/ROBOTIS-GIT/cyclo_intelligence) by ROBOTIS.
config.json ADDED
@@ -0,0 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "action_horizon": 40,
3
+ "add_pos_embed": true,
4
+ "apply_sincos_state_encoding": false,
5
+ "architectures": [
6
+ "Gr00tN1d7"
7
+ ],
8
+ "attn_dropout": 0.2,
9
+ "attn_implementation": null,
10
+ "attn_supervision_layers": null,
11
+ "attn_supervision_loss": "kl",
12
+ "attn_supervision_margin": 0.9,
13
+ "attn_supervision_weight": 0.0,
14
+ "backbone_embedding_dim": 2048,
15
+ "backbone_trainable_params_fp32": true,
16
+ "color_jitter_params": {
17
+ "brightness": 0.3,
18
+ "contrast": 0.4,
19
+ "hue": 0.08,
20
+ "saturation": 0.5
21
+ },
22
+ "crop_fraction": 0.95,
23
+ "diffusion_model_cfg": {
24
+ "attention_head_dim": 48,
25
+ "dropout": 0.2,
26
+ "final_dropout": true,
27
+ "interleave_self_attention": true,
28
+ "norm_type": "ada_norm",
29
+ "num_attention_heads": 32,
30
+ "num_layers": 32,
31
+ "output_dim": 1024,
32
+ "positional_embeddings": null
33
+ },
34
+ "dtype": "float32",
35
+ "exclude_state": false,
36
+ "formalize_language": true,
37
+ "hidden_size": 1024,
38
+ "image_crop_size": [
39
+ 230,
40
+ 230
41
+ ],
42
+ "image_target_size": [
43
+ 256,
44
+ 256
45
+ ],
46
+ "letter_box_transform": false,
47
+ "load_bf16": false,
48
+ "max_action_dim": 132,
49
+ "max_num_embodiments": 32,
50
+ "max_seq_len": 1024,
51
+ "max_state_dim": 132,
52
+ "model_dtype": "bfloat16",
53
+ "model_name": "nvidia/Cosmos-Reason2-2B",
54
+ "model_type": "Gr00tN1d7",
55
+ "noise_beta_alpha": 1.5,
56
+ "noise_beta_beta": 1.0,
57
+ "noise_s": 0.999,
58
+ "num_inference_timesteps": 4,
59
+ "num_timestep_buckets": 1000,
60
+ "random_history_crop": true,
61
+ "random_rotation_angle": 0,
62
+ "reproject_vision": false,
63
+ "rtc_ramp_rate": 6.0,
64
+ "select_layer": 16,
65
+ "shortest_image_edge": 256,
66
+ "state_dropout_prob": 0.2,
67
+ "state_gaussian_noise_std": 0.0,
68
+ "transformers_version": "4.57.3",
69
+ "ttrtc_enabled": false,
70
+ "ttrtc_max_delay_steps": 10,
71
+ "ttrtc_policy_hz": 15.0,
72
+ "ttrtc_zero_delay_probability": 0.1,
73
+ "tune_diffusion_model": true,
74
+ "tune_linear": true,
75
+ "tune_llm": false,
76
+ "tune_projector": true,
77
+ "tune_top_llm_layers": 0,
78
+ "tune_visual": false,
79
+ "tune_vlln": true,
80
+ "use_albumentations": true,
81
+ "use_alternate_vl_dit": true,
82
+ "use_flash_attention": true,
83
+ "use_future_tokens": false,
84
+ "use_mean_std": false,
85
+ "use_percentiles": true,
86
+ "use_vl_self_attention": true,
87
+ "use_vlln": true,
88
+ "vl_self_attention_cfg": {
89
+ "attention_head_dim": 64,
90
+ "dropout": 0.2,
91
+ "final_dropout": true,
92
+ "num_attention_heads": 32,
93
+ "num_layers": 4,
94
+ "positional_embeddings": null
95
+ }
96
+ }
embodiment_id.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "robocasa_panda_omron": 13,
3
+ "oxe_droid": 17,
4
+ "oxe_fractal": 18,
5
+ "oxe_language_table": 19,
6
+ "oxe_bridge": 20,
7
+ "unknown": 22,
8
+ "gr1_unified": 20,
9
+ "agibot": 26,
10
+ "sim_behavior_r1_pro": 23,
11
+ "xdof": 24,
12
+ "xdof_oss_data": 25,
13
+ "unitree_g1_full_body_with_waist_height_nav_cmd": 25,
14
+ "real_r1_pro_sharpa": 27,
15
+ "real_r1_pro_sharpa_add_view": 27,
16
+ "real_r1_pro_sharpa_relative_arm_joint": 26,
17
+ "real_r1_pro_sharpa_delta_eef": 26,
18
+ "real_r1_pro_sharpa_absolute_eef": 26,
19
+ "real_r1_pro_sharpa_meanstd": 26,
20
+ "real_r1_pro_sharpa_relative_eef": 26,
21
+ "real_r1_pro_sharpa_relative_eef_add_view": 26,
22
+ "real_r1_pro_sharpa_relative_eef_relative_hand": 26,
23
+ "real_r1_pro_sharpa_relative_eef_human": 26,
24
+ "real_r1_pro_sharpa_relative_eef_human_add_view": 26,
25
+ "real_r1_pro_sharpa_relative_eef_human_relative_hand": 26,
26
+ "real_r1_pro_sharpa_relative_eef_egodex": 26,
27
+ "real_r1_pro_sharpa_relative_eef_egodex_relative_hand": 26,
28
+ "real_r1_pro_sharpa_relative_eef_egodex_wrist_only": 26,
29
+ "real_r1_pro_sharpa_relative_eef_maxinsights": 26,
30
+ "real_r1_pro_sharpa_relative_eef_maxinsights_relative_hand": 26,
31
+ "real_r1_pro_sharpa_relative_eef_mecka": 26,
32
+ "real_r1_pro_sharpa_relative_eef_mecka_relative_hand": 26,
33
+ "real_g1_relative_eef_absolute_joints": 25,
34
+ "real_g1_relative_eef_absolute_joints_wrist_cam": 25,
35
+ "real_g1_relative_eef_relative_joints": 25,
36
+ "real_r1_pro_sharpa_relative_eef_relative_hand_relative_joint": 26,
37
+ "real_r1_pro_sharpa_relative_joint": 29,
38
+ "oxe_droid_relative_eef_relative_joint": 24,
39
+ "oxe_droid_relative_eef_relative_joint_swapped": 24,
40
+ "oxe_droid_relative_eef_relative_joint_upweight_z": 24,
41
+ "oxe_droid_relative_eef_relative_joint_upweight_z_swapped": 24,
42
+ "oxe_droid_relative_eef_relative_joint_3view": 24,
43
+ "oxe_droid_relative_eef_relative_joint_3view_swapped": 24,
44
+ "oxe_droid_relative_eef": 24,
45
+ "oxe_droid_joint_position_relative": 24,
46
+ "xdof_relative_eef_relative_joint": 27,
47
+ "xdof_relative_eef_relative_joint_subtask": 27,
48
+ "xdof_relative_eef": 27,
49
+ "xdof_relative_joint": 28,
50
+ "simpler_env_google": 0,
51
+ "simpler_env_widowx": 1,
52
+ "libero_sim": 2,
53
+ "droid_sim": 3,
54
+ "robocasa_gr1_tabletop": 10,
55
+ "new_embodiment": 10,
56
+ "unitree_g1_sonic": 11
57
+ }
experiment_cfg/conf.yaml ADDED
@@ -0,0 +1,232 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ load_config_path: null
2
+ model:
3
+ model_type: Gr00tN1d7
4
+ model_dtype: bfloat16
5
+ model_name: nvidia/Cosmos-Reason2-2B
6
+ backbone_model_type: qwen
7
+ model_revision: null
8
+ tune_top_llm_layers: 0
9
+ backbone_embedding_dim: 2048
10
+ tune_llm: false
11
+ tune_visual: false
12
+ select_layer: 12
13
+ reproject_vision: false
14
+ use_flash_attention: true
15
+ load_bf16: false
16
+ backbone_trainable_params_fp32: true
17
+ image_crop_size: null
18
+ image_target_size: null
19
+ shortest_image_edge: 256
20
+ crop_fraction: 0.9
21
+ random_rotation_angle: null
22
+ color_jitter_params: null
23
+ use_albumentations_transforms: true
24
+ letter_box_transform: true
25
+ extra_augmentation_config: null
26
+ attn_supervision_layers: null
27
+ attn_supervision_weight: 0.0
28
+ attn_supervision_loss: kl
29
+ attn_supervision_margin: 0.9
30
+ formalize_language: true
31
+ apply_sincos_state_encoding: false
32
+ use_percentiles: true
33
+ use_relative_action: true
34
+ max_state_dim: 132
35
+ max_action_dim: 132
36
+ action_horizon: 40
37
+ hidden_size: 1024
38
+ input_embedding_dim: 1536
39
+ state_history_length: 1
40
+ add_pos_embed: true
41
+ attn_dropout: 0.2
42
+ use_vlln: true
43
+ max_seq_len: 1024
44
+ use_alternate_vl_dit: true
45
+ attend_text_every_n_blocks: 2
46
+ diffusion_model_cfg:
47
+ positional_embeddings: null
48
+ num_layers: 16
49
+ num_attention_heads: 32
50
+ attention_head_dim: 48
51
+ norm_type: ada_norm
52
+ dropout: 0.2
53
+ final_dropout: true
54
+ output_dim: 1024
55
+ interleave_self_attention: true
56
+ num_inference_timesteps: 4
57
+ noise_beta_alpha: 1.5
58
+ noise_beta_beta: 1.0
59
+ noise_s: 0.999
60
+ num_timestep_buckets: 1000
61
+ ttrtc_enabled: false
62
+ ttrtc_max_delay_steps: 10
63
+ ttrtc_zero_delay_probability: 0.1
64
+ ttrtc_policy_hz: 15.0
65
+ tune_projector: true
66
+ tune_diffusion_model: true
67
+ tune_vlln: true
68
+ state_dropout_prob: 0.2
69
+ exclude_state: false
70
+ use_mean_std: false
71
+ max_num_embodiments: 32
72
+ data:
73
+ datasets:
74
+ - dataset_paths:
75
+ - /workspace/data/task000802_sh5_lerobot
76
+ embodiment_tag: new_embodiment
77
+ mix_ratio: 1.0
78
+ dataset_type: physical_embodiment
79
+ val_dataset_path: null
80
+ modality_configs:
81
+ new_embodiment:
82
+ video:
83
+ delta_indices:
84
+ - 0
85
+ modality_keys:
86
+ - cam_left_head
87
+ - cam_left_wrist
88
+ - cam_right_wrist
89
+ sin_cos_embedding_keys: null
90
+ mean_std_embedding_keys: null
91
+ action_configs: null
92
+ state:
93
+ delta_indices:
94
+ - 0
95
+ modality_keys:
96
+ - arm_left
97
+ - arm_right
98
+ - hand_left
99
+ - hand_right
100
+ sin_cos_embedding_keys: null
101
+ mean_std_embedding_keys: null
102
+ action_configs: null
103
+ action:
104
+ delta_indices:
105
+ - 0
106
+ - 1
107
+ - 2
108
+ - 3
109
+ - 4
110
+ - 5
111
+ - 6
112
+ - 7
113
+ - 8
114
+ - 9
115
+ - 10
116
+ - 11
117
+ - 12
118
+ - 13
119
+ - 14
120
+ - 15
121
+ modality_keys:
122
+ - arm_left
123
+ - arm_right
124
+ - hand_left
125
+ - hand_right
126
+ sin_cos_embedding_keys: null
127
+ mean_std_embedding_keys: null
128
+ action_configs:
129
+ - rep: ABSOLUTE
130
+ type: NON_EEF
131
+ format: DEFAULT
132
+ state_key: null
133
+ - rep: ABSOLUTE
134
+ type: NON_EEF
135
+ format: DEFAULT
136
+ state_key: null
137
+ - rep: ABSOLUTE
138
+ type: NON_EEF
139
+ format: DEFAULT
140
+ state_key: null
141
+ - rep: ABSOLUTE
142
+ type: NON_EEF
143
+ format: DEFAULT
144
+ state_key: null
145
+ language:
146
+ delta_indices:
147
+ - 0
148
+ modality_keys:
149
+ - sub_task
150
+ sin_cos_embedding_keys: null
151
+ mean_std_embedding_keys: null
152
+ action_configs: null
153
+ download_cache: false
154
+ shard_size: 1024
155
+ episode_sampling_rate: 0.1
156
+ num_shards_per_epoch: 100000
157
+ ds_weights_alpha: null
158
+ override_pretraining_statistics: true
159
+ mode: single_turn
160
+ random_chop: 0.0
161
+ mock_dataset_mode: false
162
+ shuffle: true
163
+ seed: 42
164
+ multiprocessing_context: fork
165
+ allow_padding: false
166
+ subsample_ratio: 1.0
167
+ image_crop_size:
168
+ - 244
169
+ - 244
170
+ image_target_size:
171
+ - 224
172
+ - 224
173
+ training:
174
+ output_dir: /workspace/model/groot/RobotisSW
175
+ experiment_name: task000802_sh5_groot_n17_standard_100k_v1
176
+ max_steps: 100000
177
+ global_batch_size: 4
178
+ per_gpu_batch_size: null
179
+ gradient_accumulation_steps: 6
180
+ learning_rate: 0.0001
181
+ lr_scheduler_type: cosine
182
+ weight_decay: 1.0e-05
183
+ warmup_ratio: 0.05
184
+ warmup_steps: 0
185
+ max_grad_norm: 1.0
186
+ optim: adamw_bnb_8bit
187
+ start_from_checkpoint: nvidia/GR00T-N1.7-3B
188
+ skip_weight_loading: false
189
+ tf32: true
190
+ fp16: false
191
+ bf16: true
192
+ eval_bf16: true
193
+ logging_steps: 10
194
+ save_steps: 10000
195
+ save_total_limit: 3
196
+ save_vl_model: false
197
+ save_only_model: false
198
+ resume_from_checkpoint: false
199
+ upload_checkpoints: false
200
+ upload_every: 1000
201
+ upload_last_n_checkpoints: 5
202
+ max_concurrent_uploads: 2
203
+ eval_strategy: 'no'
204
+ eval_steps: 500
205
+ eval_set_split_ratio: 0.1
206
+ eval_batch_size: 2
207
+ save_best_eval_metric_name: ''
208
+ save_best_eval_metric_greater_is_better: true
209
+ deepspeed_stage: 2
210
+ gradient_checkpointing: false
211
+ transformers_trust_remote_code: true
212
+ transformers_local_files_only: false
213
+ transformers_cache_dir: null
214
+ transformers_access_token: null
215
+ use_ddp: false
216
+ ddp_bucket_cap_mb: 100
217
+ num_gpus: 1
218
+ dataloader_num_workers: 4
219
+ remove_unused_columns: false
220
+ use_wandb: false
221
+ wandb_project: finetune-gr00t-n1d7
222
+ enable_profiling: false
223
+ max_retries: 3
224
+ assert_loss_less_than: null
225
+ add_rl_callback: false
226
+ enable_open_loop_eval: false
227
+ open_loop_eval_traj_ids:
228
+ - 0
229
+ open_loop_eval_steps_per_traj: 100
230
+ open_loop_eval_plot_indices: null
231
+ max_steps: 100000
232
+ save_steps: 10000
experiment_cfg/config.yaml ADDED
@@ -0,0 +1,230 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ load_config_path: null
2
+ model:
3
+ model_type: Gr00tN1d7
4
+ model_dtype: bfloat16
5
+ model_name: nvidia/Cosmos-Reason2-2B
6
+ backbone_model_type: qwen
7
+ model_revision: null
8
+ tune_top_llm_layers: 0
9
+ backbone_embedding_dim: 2048
10
+ tune_llm: false
11
+ tune_visual: false
12
+ select_layer: 12
13
+ reproject_vision: false
14
+ use_flash_attention: true
15
+ load_bf16: false
16
+ backbone_trainable_params_fp32: true
17
+ image_crop_size: null
18
+ image_target_size: null
19
+ shortest_image_edge: 256
20
+ crop_fraction: 0.9
21
+ random_rotation_angle: null
22
+ color_jitter_params: null
23
+ use_albumentations_transforms: true
24
+ letter_box_transform: true
25
+ extra_augmentation_config: null
26
+ attn_supervision_layers: null
27
+ attn_supervision_weight: 0.0
28
+ attn_supervision_loss: kl
29
+ attn_supervision_margin: 0.9
30
+ formalize_language: true
31
+ apply_sincos_state_encoding: false
32
+ use_percentiles: true
33
+ use_relative_action: true
34
+ max_state_dim: 132
35
+ max_action_dim: 132
36
+ action_horizon: 40
37
+ hidden_size: 1024
38
+ input_embedding_dim: 1536
39
+ state_history_length: 1
40
+ add_pos_embed: true
41
+ attn_dropout: 0.2
42
+ use_vlln: true
43
+ max_seq_len: 1024
44
+ use_alternate_vl_dit: true
45
+ attend_text_every_n_blocks: 2
46
+ diffusion_model_cfg:
47
+ positional_embeddings: null
48
+ num_layers: 16
49
+ num_attention_heads: 32
50
+ attention_head_dim: 48
51
+ norm_type: ada_norm
52
+ dropout: 0.2
53
+ final_dropout: true
54
+ output_dim: 1024
55
+ interleave_self_attention: true
56
+ num_inference_timesteps: 4
57
+ noise_beta_alpha: 1.5
58
+ noise_beta_beta: 1.0
59
+ noise_s: 0.999
60
+ num_timestep_buckets: 1000
61
+ ttrtc_enabled: false
62
+ ttrtc_max_delay_steps: 10
63
+ ttrtc_zero_delay_probability: 0.1
64
+ ttrtc_policy_hz: 15.0
65
+ tune_projector: true
66
+ tune_diffusion_model: true
67
+ tune_vlln: true
68
+ state_dropout_prob: 0.2
69
+ exclude_state: false
70
+ use_mean_std: false
71
+ max_num_embodiments: 32
72
+ data:
73
+ datasets:
74
+ - dataset_paths:
75
+ - /workspace/data/task000802_sh5_lerobot
76
+ embodiment_tag: new_embodiment
77
+ mix_ratio: 1.0
78
+ dataset_type: physical_embodiment
79
+ val_dataset_path: null
80
+ modality_configs:
81
+ new_embodiment:
82
+ video:
83
+ delta_indices:
84
+ - 0
85
+ modality_keys:
86
+ - cam_left_head
87
+ - cam_left_wrist
88
+ - cam_right_wrist
89
+ sin_cos_embedding_keys: null
90
+ mean_std_embedding_keys: null
91
+ action_configs: null
92
+ state:
93
+ delta_indices:
94
+ - 0
95
+ modality_keys:
96
+ - arm_left
97
+ - arm_right
98
+ - hand_left
99
+ - hand_right
100
+ sin_cos_embedding_keys: null
101
+ mean_std_embedding_keys: null
102
+ action_configs: null
103
+ action:
104
+ delta_indices:
105
+ - 0
106
+ - 1
107
+ - 2
108
+ - 3
109
+ - 4
110
+ - 5
111
+ - 6
112
+ - 7
113
+ - 8
114
+ - 9
115
+ - 10
116
+ - 11
117
+ - 12
118
+ - 13
119
+ - 14
120
+ - 15
121
+ modality_keys:
122
+ - arm_left
123
+ - arm_right
124
+ - hand_left
125
+ - hand_right
126
+ sin_cos_embedding_keys: null
127
+ mean_std_embedding_keys: null
128
+ action_configs:
129
+ - rep: absolute
130
+ type: non_eef
131
+ format: default
132
+ state_key: null
133
+ - rep: absolute
134
+ type: non_eef
135
+ format: default
136
+ state_key: null
137
+ - rep: absolute
138
+ type: non_eef
139
+ format: default
140
+ state_key: null
141
+ - rep: absolute
142
+ type: non_eef
143
+ format: default
144
+ state_key: null
145
+ language:
146
+ delta_indices:
147
+ - 0
148
+ modality_keys:
149
+ - sub_task
150
+ sin_cos_embedding_keys: null
151
+ mean_std_embedding_keys: null
152
+ action_configs: null
153
+ download_cache: false
154
+ shard_size: 1024
155
+ episode_sampling_rate: 0.1
156
+ num_shards_per_epoch: 100000
157
+ ds_weights_alpha: null
158
+ override_pretraining_statistics: true
159
+ mode: single_turn
160
+ random_chop: 0.0
161
+ mock_dataset_mode: false
162
+ shuffle: true
163
+ seed: 42
164
+ multiprocessing_context: fork
165
+ allow_padding: false
166
+ subsample_ratio: 1.0
167
+ image_crop_size:
168
+ - 244
169
+ - 244
170
+ image_target_size:
171
+ - 224
172
+ - 224
173
+ training:
174
+ output_dir: /workspace/model/groot/RobotisSW
175
+ experiment_name: task000802_sh5_groot_n17_standard_100k_v1
176
+ max_steps: 100000
177
+ global_batch_size: 4
178
+ per_gpu_batch_size: null
179
+ gradient_accumulation_steps: 6
180
+ learning_rate: 0.0001
181
+ lr_scheduler_type: cosine
182
+ weight_decay: 1.0e-05
183
+ warmup_ratio: 0.05
184
+ warmup_steps: 0
185
+ max_grad_norm: 1.0
186
+ optim: adamw_bnb_8bit
187
+ start_from_checkpoint: nvidia/GR00T-N1.7-3B
188
+ skip_weight_loading: false
189
+ tf32: true
190
+ fp16: false
191
+ bf16: true
192
+ eval_bf16: true
193
+ logging_steps: 10
194
+ save_steps: 10000
195
+ save_total_limit: 3
196
+ save_vl_model: false
197
+ save_only_model: false
198
+ resume_from_checkpoint: false
199
+ upload_checkpoints: false
200
+ upload_every: 1000
201
+ upload_last_n_checkpoints: 5
202
+ max_concurrent_uploads: 2
203
+ eval_strategy: 'no'
204
+ eval_steps: 500
205
+ eval_set_split_ratio: 0.1
206
+ eval_batch_size: 2
207
+ save_best_eval_metric_name: ''
208
+ save_best_eval_metric_greater_is_better: true
209
+ deepspeed_stage: 2
210
+ gradient_checkpointing: false
211
+ transformers_trust_remote_code: true
212
+ transformers_local_files_only: false
213
+ transformers_cache_dir: null
214
+ transformers_access_token: null
215
+ use_ddp: false
216
+ ddp_bucket_cap_mb: 100
217
+ num_gpus: 1
218
+ dataloader_num_workers: 4
219
+ remove_unused_columns: false
220
+ use_wandb: false
221
+ wandb_project: finetune-gr00t-n1d7
222
+ enable_profiling: false
223
+ max_retries: 3
224
+ assert_loss_less_than: null
225
+ add_rl_callback: false
226
+ enable_open_loop_eval: false
227
+ open_loop_eval_traj_ids:
228
+ - 0
229
+ open_loop_eval_steps_per_traj: 100
230
+ open_loop_eval_plot_indices: null
experiment_cfg/dataset_statistics.json ADDED
@@ -0,0 +1,769 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "new_embodiment": {
3
+ "state": {
4
+ "arm_left": {
5
+ "min": [
6
+ -0.13344433903694153,
7
+ -1.1984225238848012e-05,
8
+ -0.1426841765642166,
9
+ -2.6476387977600098,
10
+ -0.2070874124765396,
11
+ 0.8743570446968079,
12
+ -0.1196235716342926
13
+ ],
14
+ "max": [
15
+ 0.08895890414714813,
16
+ 1.1984225238848012e-05,
17
+ -0.052143361419439316,
18
+ -2.574007749557495,
19
+ -0.04140549525618553,
20
+ 1.0929492712020874,
21
+ -0.009163624607026577
22
+ ],
23
+ "mean": [
24
+ -0.05443303659558296,
25
+ 3.178290963745667e-10,
26
+ -0.10623468458652496,
27
+ -2.609933614730835,
28
+ -0.19235453009605408,
29
+ 1.060228705406189,
30
+ -0.11073947697877884
31
+ ],
32
+ "std": [
33
+ 0.056636907160282135,
34
+ 1.8514923283419193e-07,
35
+ 0.013435847125947442,
36
+ 0.01699644513427548,
37
+ 0.032947737723588943,
38
+ 0.07274181395769082,
39
+ 0.028182165697216988
40
+ ],
41
+ "q01": [
42
+ -0.13344433903694153,
43
+ 0.0,
44
+ -0.1426602154970169,
45
+ -2.6476387977600098,
46
+ -0.2070874124765396,
47
+ 0.8743570446968079,
48
+ -0.1196235716342926
49
+ ],
50
+ "q99": [
51
+ 0.08895890414714813,
52
+ 0.0,
53
+ -0.05215534567832947,
54
+ -2.574007749557495,
55
+ -0.041417479515075684,
56
+ 1.0929492712020874,
57
+ -0.009163624607026577
58
+ ]
59
+ },
60
+ "arm_right": {
61
+ "min": [
62
+ -0.5563796162605286,
63
+ -0.6961036920547485,
64
+ -0.5945613384246826,
65
+ -2.919656753540039,
66
+ -0.6392265558242798,
67
+ -1.1935688257217407,
68
+ -0.5612978935241699
69
+ ],
70
+ "max": [
71
+ 0.37515416741371155,
72
+ 0.00016777914424892515,
73
+ 0.5537310838699341,
74
+ -1.2355976104736328,
75
+ 1.94078528881073,
76
+ 1.5700054168701172,
77
+ 1.39189875125885
78
+ ],
79
+ "mean": [
80
+ -0.12428255379199982,
81
+ -0.0766887292265892,
82
+ 0.02543916180729866,
83
+ -2.2738661766052246,
84
+ 0.6260746121406555,
85
+ 0.6683403253555298,
86
+ 0.39729464054107666
87
+ ],
88
+ "std": [
89
+ 0.14193488657474518,
90
+ 0.1045134961605072,
91
+ 0.15955595672130585,
92
+ 0.36540859937667847,
93
+ 0.5633009076118469,
94
+ 0.5089101791381836,
95
+ 0.37172114849090576
96
+ ],
97
+ "q01": [
98
+ -0.4798843264579773,
99
+ -0.5507695531845093,
100
+ -0.33662680387496946,
101
+ -2.7594635486602783,
102
+ -0.34502583742141724,
103
+ -0.9316397142410279,
104
+ -0.3560600030422211
105
+ ],
106
+ "q99": [
107
+ 0.19787920355796831,
108
+ 4.793690095539205e-05,
109
+ 0.38315819978713994,
110
+ -1.4236809158325194,
111
+ 1.660749912261963,
112
+ 1.3654782629013063,
113
+ 1.1318317651748657
114
+ ]
115
+ },
116
+ "hand_left": {
117
+ "min": [
118
+ -0.7731263041496277,
119
+ -3.1461946964263916,
120
+ -3.1247189044952393,
121
+ -0.251572847366333,
122
+ -0.1871456503868103,
123
+ 1.1382137537002563,
124
+ 0.5875146389007568,
125
+ 0.7439807057380676,
126
+ -0.10124272853136063,
127
+ 1.274738073348999,
128
+ 0.5737088322639465,
129
+ 0.6442719101905823,
130
+ 0.01073786523193121,
131
+ 1.1550875902175903,
132
+ 0.6059224009513855,
133
+ 0.7133010625839233,
134
+ -0.11351457983255386,
135
+ 0.977145791053772,
136
+ 0.5675728917121887,
137
+ 0.849825382232666
138
+ ],
139
+ "max": [
140
+ 0.09357282519340515,
141
+ 1.572330355644226,
142
+ 0.5154175162315369,
143
+ 38.262081146240234,
144
+ -0.15493206679821014,
145
+ 1.2701361179351807,
146
+ 0.6458058953285217,
147
+ 0.7593204975128174,
148
+ -0.07209710031747818,
149
+ 1.352971076965332,
150
+ 0.6135923266410828,
151
+ 0.6810874938964844,
152
+ 0.04448544234037399,
153
+ 1.2394565343856812,
154
+ 0.6458058953285217,
155
+ 0.7347767949104309,
156
+ 0.0951068103313446,
157
+ 1.0753204822540283,
158
+ 0.630466103553772,
159
+ 0.8774369955062866
160
+ ],
161
+ "mean": [
162
+ -0.6869894862174988,
163
+ 1.5081450939178467,
164
+ -0.7967520356178284,
165
+ -0.21610599756240845,
166
+ -0.1686253547668457,
167
+ 1.1971584558486938,
168
+ 0.6194696426391602,
169
+ 0.7555567622184753,
170
+ -0.08980464190244675,
171
+ 1.301042079925537,
172
+ 0.5898573398590088,
173
+ 0.6669216752052307,
174
+ 0.02541246823966503,
175
+ 1.1902872323989868,
176
+ 0.6249839663505554,
177
+ 0.7263347506523132,
178
+ 0.022257905453443527,
179
+ 1.009364366531372,
180
+ 0.6186327934265137,
181
+ 0.8608976602554321
182
+ ],
183
+ "std": [
184
+ 0.045401960611343384,
185
+ 0.03936194628477026,
186
+ 0.6181727647781372,
187
+ 0.27919796109199524,
188
+ 0.00903857033699746,
189
+ 0.03430493175983429,
190
+ 0.011666439473628998,
191
+ 0.002154000336303489,
192
+ 0.007955420762300491,
193
+ 0.026035197079180606,
194
+ 0.012292167171836886,
195
+ 0.01174754556268509,
196
+ 0.008417760021984577,
197
+ 0.029953999444841353,
198
+ 0.0118643119931221,
199
+ 0.006562825292347804,
200
+ 0.06818024814128876,
201
+ 0.02842798084020517,
202
+ 0.005368546117095827,
203
+ 0.008142561651769655
204
+ ],
205
+ "q01": [
206
+ -0.7731263041496277,
207
+ 1.492563247680664,
208
+ -1.162757396697998,
209
+ -0.25003886222839355,
210
+ -0.18561168015003204,
211
+ 1.1596894264221191,
212
+ 0.5997864603996277,
213
+ 0.7470486164093018,
214
+ -0.09970875084400177,
215
+ 1.2762720584869385,
216
+ 0.5737088322639465,
217
+ 0.6442719101905823,
218
+ 0.013805827125906944,
219
+ 1.1596894264221191,
220
+ 0.6105243563652039,
221
+ 0.7133010625839233,
222
+ -0.11198059469461441,
223
+ 0.983281672000885,
224
+ 0.6059224009513855,
225
+ 0.849825382232666
226
+ ],
227
+ "q99": [
228
+ -0.6243301630020142,
229
+ 1.5692623853683472,
230
+ 0.5138835906982422,
231
+ -0.20248545706272125,
232
+ -0.15493206679821014,
233
+ 1.2701361179351807,
234
+ 0.6442719101905823,
235
+ 0.7593204975128174,
236
+ -0.07209710031747818,
237
+ 1.3514370918273926,
238
+ 0.6105243563652039,
239
+ 0.676485538482666,
240
+ 0.04448544234037399,
241
+ 1.2379225492477417,
242
+ 0.6442719101905823,
243
+ 0.731708824634552,
244
+ 0.0951068103313446,
245
+ 1.06918466091156,
246
+ 0.6258641481399536,
247
+ 0.8774369955062866
248
+ ]
249
+ },
250
+ "hand_right": {
251
+ "min": [
252
+ -0.02454369328916073,
253
+ -1.7027186155319214,
254
+ -0.4724660813808441,
255
+ -0.01840776950120926,
256
+ 0.01073786523193121,
257
+ -0.02761165425181389,
258
+ -0.019941750913858414,
259
+ -0.01840776950120926,
260
+ -0.02761165425181389,
261
+ -0.006135923322290182,
262
+ -0.01073786523193121,
263
+ -0.015339807607233524,
264
+ -0.06749515235424042,
265
+ -0.019941750913858414,
266
+ -0.016873788088560104,
267
+ -0.026077672839164734,
268
+ -0.1564660370349884,
269
+ -0.03988350182771683,
270
+ -0.02147573046386242,
271
+ -0.029145635664463043
272
+ ],
273
+ "max": [
274
+ 0.777728259563446,
275
+ -1.414330244064331,
276
+ 1.116737961769104,
277
+ 0.23316508531570435,
278
+ 0.42031073570251465,
279
+ 40.84070587158203,
280
+ 41.233402252197266,
281
+ 41.626102447509766,
282
+ 0.1257864236831665,
283
+ 1.336097240447998,
284
+ 0.5967185497283936,
285
+ 0.6841554045677185,
286
+ 0.03067961521446705,
287
+ 1.2302526235580444,
288
+ 0.6473398804664612,
289
+ 0.7593204975128174,
290
+ 0.01840776950120926,
291
+ 0.9970875382423401,
292
+ 0.6273981332778931,
293
+ 0.9019806981086731
294
+ ],
295
+ "mean": [
296
+ 0.39105281233787537,
297
+ -1.5267596244812012,
298
+ 0.4859321117401123,
299
+ 0.08365841209888458,
300
+ 0.10027659684419632,
301
+ 0.6473800539970398,
302
+ 0.33088418841362,
303
+ 0.4408280849456787,
304
+ 0.056424129754304886,
305
+ 0.7307093143463135,
306
+ 0.309431791305542,
307
+ 0.3837047815322876,
308
+ -0.006917295977473259,
309
+ 0.6774591207504272,
310
+ 0.34364423155784607,
311
+ 0.42524340748786926,
312
+ -0.061637263745069504,
313
+ 0.5484294891357422,
314
+ 0.34025245904922485,
315
+ 0.5194066762924194
316
+ ],
317
+ "std": [
318
+ 0.22203458845615387,
319
+ 0.044938161969182405,
320
+ 0.2841412425041199,
321
+ 0.057933323085308075,
322
+ 0.038640838116407394,
323
+ 0.38714873790740967,
324
+ 0.2698653042316437,
325
+ 0.3145829439163208,
326
+ 0.023706894367933273,
327
+ 0.36628687381744385,
328
+ 0.14706483483314514,
329
+ 0.19983762502670288,
330
+ 0.012456710450351238,
331
+ 0.3368881940841675,
332
+ 0.1683698147535324,
333
+ 0.22269974648952484,
334
+ 0.03867242485284805,
335
+ 0.2834411859512329,
336
+ 0.17644259333610535,
337
+ 0.2786138355731964
338
+ ],
339
+ "q01": [
340
+ 0.08130098134279251,
341
+ -1.6643691062927246,
342
+ -0.16971963882446287,
343
+ -0.0015339808305725455,
344
+ 0.03681553900241852,
345
+ 0.16566991806030273,
346
+ 0.08436894416809082,
347
+ 0.10124272853136063,
348
+ 0.007669903803616762,
349
+ 0.17947575449943542,
350
+ 0.0782330185174942,
351
+ 0.08590292185544968,
352
+ -0.041417479515075684,
353
+ 0.17180584371089935,
354
+ 0.08436894416809082,
355
+ 0.09357282519340515,
356
+ -0.14572817087173462,
357
+ 0.12885437905788422,
358
+ 0.07976700365543365,
359
+ 0.11965049803256989
360
+ ],
361
+ "q99": [
362
+ 0.7593204975128174,
363
+ -1.4649516344070435,
364
+ 1.1075341701507568,
365
+ 0.22242721915245056,
366
+ 0.17947575449943542,
367
+ 1.1995729207992554,
368
+ 0.6320000886917114,
369
+ 0.7746602892875671,
370
+ 0.11351457983255386,
371
+ 1.2946797609329224,
372
+ 0.5568350553512573,
373
+ 0.6580777764320374,
374
+ 0.013805827125906944,
375
+ 1.2041748762130737,
376
+ 0.6151263117790222,
377
+ 0.7271068692207336,
378
+ 0.0015339808305725455,
379
+ 0.9924855828285217,
380
+ 0.6181942820549011,
381
+ 0.8958447575569153
382
+ ]
383
+ }
384
+ },
385
+ "action": {
386
+ "arm_left": {
387
+ "min": [
388
+ -0.13345633447170258,
389
+ 0.0,
390
+ -0.1426602154970169,
391
+ -2.647650957107544,
392
+ -0.2070874124765396,
393
+ 0.8743690252304077,
394
+ -0.1196235716342926
395
+ ],
396
+ "max": [
397
+ 0.08897088468074799,
398
+ 1.1984225238848012e-05,
399
+ -0.05215534567832947,
400
+ -2.57401967048645,
401
+ -0.041417479515075684,
402
+ 1.0929492712020874,
403
+ -0.00920388475060463
404
+ ],
405
+ "mean": [
406
+ -0.05443453788757324,
407
+ 4.767437555841525e-10,
408
+ -0.10623478144407272,
409
+ -2.609933614730835,
410
+ -0.19235457479953766,
411
+ 1.060228705406189,
412
+ -0.11074069887399673
413
+ ],
414
+ "std": [
415
+ 0.056641533970832825,
416
+ 7.558458747780605e-08,
417
+ 0.013436375185847282,
418
+ 0.01699821278453159,
419
+ 0.03294767066836357,
420
+ 0.07274182885885201,
421
+ 0.028173090890049934
422
+ ],
423
+ "q01": [
424
+ -0.13345633447170258,
425
+ 0.0,
426
+ -0.1426602154970169,
427
+ -2.647650957107544,
428
+ -0.2070874124765396,
429
+ 0.8743690252304077,
430
+ -0.1196235716342926
431
+ ],
432
+ "q99": [
433
+ 0.08897088468074799,
434
+ 0.0,
435
+ -0.05215534567832947,
436
+ -2.57401967048645,
437
+ -0.041417479515075684,
438
+ 1.0929492712020874,
439
+ -0.00920388475060463
440
+ ]
441
+ },
442
+ "arm_right": {
443
+ "min": [
444
+ -0.5568350553512573,
445
+ -0.6979612708091736,
446
+ -0.598252534866333,
447
+ -2.9176313877105713,
448
+ -0.6419709324836731,
449
+ -1.1942040920257568,
450
+ -0.5614369511604309
451
+ ],
452
+ "max": [
453
+ 0.37582528591156006,
454
+ 0.0,
455
+ 0.5537670850753784,
456
+ -1.2348545789718628,
457
+ 1.9420197010040283,
458
+ 1.5700000524520874,
459
+ 1.3920875787734985
460
+ ],
461
+ "mean": [
462
+ -0.12424712628126144,
463
+ -0.07647409290075302,
464
+ 0.025355221703648567,
465
+ -2.273934841156006,
466
+ 0.6256620287895203,
467
+ 0.6685455441474915,
468
+ 0.39710843563079834
469
+ ],
470
+ "std": [
471
+ 0.14228683710098267,
472
+ 0.1047787293791771,
473
+ 0.16147425770759583,
474
+ 0.36621159315109253,
475
+ 0.5653416514396667,
476
+ 0.5099804401397705,
477
+ 0.37255698442459106
478
+ ],
479
+ "q01": [
480
+ -0.480135977268219,
481
+ -0.552233099937439,
482
+ -0.34514567255973816,
483
+ -2.7611653804779053,
484
+ -0.3543495535850525,
485
+ -0.9341943264007568,
486
+ -0.365854412317276
487
+ ],
488
+ "q99": [
489
+ 0.1979724884033208,
490
+ 0.0,
491
+ 0.3834952116012573,
492
+ -1.422184247970574,
493
+ 1.6613011360168457,
494
+ 1.3690778017044067,
495
+ 1.1320778131484985
496
+ ]
497
+ },
498
+ "hand_left": {
499
+ "min": [
500
+ -0.7731263041496277,
501
+ -3.1461946964263916,
502
+ -3.1247189044952393,
503
+ -0.251572847366333,
504
+ -0.1871456503868103,
505
+ 1.1382137537002563,
506
+ 0.5875146389007568,
507
+ 0.7439807057380676,
508
+ -0.10124272853136063,
509
+ 1.274738073348999,
510
+ 0.5737088322639465,
511
+ 0.6442719101905823,
512
+ 0.01073786523193121,
513
+ 1.1550875902175903,
514
+ 0.6059224009513855,
515
+ 0.7133010625839233,
516
+ -0.11198059469461441,
517
+ 0.977145791053772,
518
+ 0.5660389065742493,
519
+ 0.849825382232666
520
+ ],
521
+ "max": [
522
+ 0.09357282519340515,
523
+ 1.572330355644226,
524
+ 0.5154175162315369,
525
+ 38.262081146240234,
526
+ -0.15493206679821014,
527
+ 1.2701361179351807,
528
+ 0.6442719101905823,
529
+ 0.7593204975128174,
530
+ -0.07209710031747818,
531
+ 1.3545050621032715,
532
+ 0.6135923266410828,
533
+ 0.6810874938964844,
534
+ 0.04448544234037399,
535
+ 1.2394565343856812,
536
+ 0.6458058953285217,
537
+ 0.7332428097724915,
538
+ 0.0951068103313446,
539
+ 1.0753204822540283,
540
+ 0.630466103553772,
541
+ 0.8774369955062866
542
+ ],
543
+ "mean": [
544
+ -0.6870015859603882,
545
+ 1.5082069635391235,
546
+ -0.7967050075531006,
547
+ -0.21661075949668884,
548
+ -0.1686253398656845,
549
+ 1.1971582174301147,
550
+ 0.6194697022438049,
551
+ 0.7555566430091858,
552
+ -0.08980489522218704,
553
+ 1.3010419607162476,
554
+ 0.5898569226264954,
555
+ 0.6669216156005859,
556
+ 0.025412453338503838,
557
+ 1.190285563468933,
558
+ 0.6249831914901733,
559
+ 0.7263347506523132,
560
+ 0.02225753478705883,
561
+ 1.0093637704849243,
562
+ 0.6186316013336182,
563
+ 0.8608974814414978
564
+ ],
565
+ "std": [
566
+ 0.04531852528452858,
567
+ 0.035534288734200545,
568
+ 0.61813884973526,
569
+ 0.24231840670108795,
570
+ 0.009038486517965745,
571
+ 0.03430548310279846,
572
+ 0.01166701223701235,
573
+ 0.0021541994065038257,
574
+ 0.007955417968332714,
575
+ 0.026035418733952764,
576
+ 0.012292096391321216,
577
+ 0.011747491545976057,
578
+ 0.008418328128755093,
579
+ 0.029954729601740778,
580
+ 0.011863167397676984,
581
+ 0.006562782917170511,
582
+ 0.06817995756864548,
583
+ 0.028427625074984048,
584
+ 0.005368172191083351,
585
+ 0.00814257096498901
586
+ ],
587
+ "q01": [
588
+ -0.7731263041496277,
589
+ 1.492563247680664,
590
+ -1.162757396697998,
591
+ -0.25003886222839355,
592
+ -0.18561168015003204,
593
+ 1.1596894264221191,
594
+ 0.5997864603996277,
595
+ 0.7470486164093018,
596
+ -0.09970875084400177,
597
+ 1.2762720584869385,
598
+ 0.5737088322639465,
599
+ 0.6442719101905823,
600
+ 0.013805827125906944,
601
+ 1.1596894264221191,
602
+ 0.6105243563652039,
603
+ 0.7133010625839233,
604
+ -0.11198059469461441,
605
+ 0.983281672000885,
606
+ 0.6059224009513855,
607
+ 0.849825382232666
608
+ ],
609
+ "q99": [
610
+ -0.6243301630020142,
611
+ 1.5692623853683472,
612
+ 0.5138835906982422,
613
+ -0.20248545706272125,
614
+ -0.15493206679821014,
615
+ 1.2701361179351807,
616
+ 0.6442719101905823,
617
+ 0.7593204975128174,
618
+ -0.07209710031747818,
619
+ 1.3514370918273926,
620
+ 0.6105243563652039,
621
+ 0.676485538482666,
622
+ 0.04448544234037399,
623
+ 1.2379225492477417,
624
+ 0.6442719101905823,
625
+ 0.731708824634552,
626
+ 0.0951068103313446,
627
+ 1.06918466091156,
628
+ 0.6258641481399536,
629
+ 0.8774369955062866
630
+ ]
631
+ },
632
+ "hand_right": {
633
+ "min": [
634
+ 0.041871316730976105,
635
+ -1.7027186155319214,
636
+ -0.4724660813808441,
637
+ 0.01439326535910368,
638
+ 0.009813589975237846,
639
+ 0.07850871980190277,
640
+ 0.041217077523469925,
641
+ 0.05037643015384674,
642
+ 0.00327119673602283,
643
+ 0.08570535480976105,
644
+ 0.039908599108457565,
645
+ 0.04448827728629112,
646
+ -0.05000000074505806,
647
+ 0.07981719821691513,
648
+ 0.041871316730976105,
649
+ 0.0484137125313282,
650
+ -0.15000000596046448,
651
+ 0.06411545723676682,
652
+ 0.039908599108457565,
653
+ 0.05757306143641472
654
+ ],
655
+ "max": [
656
+ 0.7685243487358093,
657
+ -1.477223515510559,
658
+ 1.1121361255645752,
659
+ 0.2316311001777649,
660
+ 0.17947575449943542,
661
+ 1.2000000476837158,
662
+ 0.6299999952316284,
663
+ 0.7699999809265137,
664
+ 0.1089126393198967,
665
+ 1.309999942779541,
666
+ 0.6100000143051147,
667
+ 0.6800000071525574,
668
+ 0.00920388475060463,
669
+ 1.228718638420105,
670
+ 0.6399999856948853,
671
+ 0.7400000095367432,
672
+ -0.009813589975237846,
673
+ 0.9955535531044006,
674
+ 0.6181942820549011,
675
+ 0.8927768468856812
676
+ ],
677
+ "mean": [
678
+ 0.39305078983306885,
679
+ -1.503967523574829,
680
+ 0.6515370011329651,
681
+ 0.13150832056999207,
682
+ 0.09349370002746582,
683
+ 0.7223532199859619,
684
+ 0.3799966275691986,
685
+ 0.4650123417377472,
686
+ 0.036148712038993835,
687
+ 0.791204571723938,
688
+ 0.3617909848690033,
689
+ 0.40839266777038574,
690
+ -0.025904934853315353,
691
+ 0.7340635657310486,
692
+ 0.3828246593475342,
693
+ 0.445637583732605,
694
+ -0.08410117030143738,
695
+ 0.5932419896125793,
696
+ 0.36932283639907837,
697
+ 0.532431960105896
698
+ ],
699
+ "std": [
700
+ 0.21111871302127838,
701
+ 0.018326751887796793,
702
+ 0.3620738089084625,
703
+ 0.07063298672437668,
704
+ 0.05118520185351372,
705
+ 0.3803084194660187,
706
+ 0.20047608017921448,
707
+ 0.24517615139484406,
708
+ 0.025977257639169693,
709
+ 0.4173443019390106,
710
+ 0.18846257030963898,
711
+ 0.21425677835941315,
712
+ 0.015481325797736645,
713
+ 0.3860543668270111,
714
+ 0.20042333006858826,
715
+ 0.2343057543039322,
716
+ 0.04414879530668259,
717
+ 0.31345513463020325,
718
+ 0.19528095424175262,
719
+ 0.2810819447040558
720
+ ],
721
+ "q01": [
722
+ 0.09406757354736328,
723
+ -1.604543924331665,
724
+ 0.14311747252941132,
725
+ 0.030479609966278076,
726
+ 0.02204708755016327,
727
+ 0.17637670040130615,
728
+ 0.09259776771068573,
729
+ 0.11317504942417145,
730
+ 0.007349029183387756,
731
+ 0.19254456460475922,
732
+ 0.08965815603733063,
733
+ 0.09994679689407349,
734
+ -0.05000000074505806,
735
+ 0.17931631207466125,
736
+ 0.09406757354736328,
737
+ 0.1087656319141388,
738
+ -0.15000000596046448,
739
+ 0.14404097199440002,
740
+ 0.08965815603733063,
741
+ 0.1293429136276245
742
+ ],
743
+ "q99": [
744
+ 0.6795535087585449,
745
+ -1.5,
746
+ 1.104466199874878,
747
+ 0.22242721915245056,
748
+ 0.17640778422355652,
749
+ 1.2000000476837158,
750
+ 0.6299999952316284,
751
+ 0.7699999809265137,
752
+ 0.10124272853136063,
753
+ 1.309999942779541,
754
+ 0.6100000143051147,
755
+ 0.6800000071525574,
756
+ -0.006927184294909239,
757
+ 1.2200000286102295,
758
+ 0.6399999856948853,
759
+ 0.7400000095367432,
760
+ -0.02204708755016327,
761
+ 0.9800000190734863,
762
+ 0.6105243563652039,
763
+ 0.8799999952316284
764
+ ]
765
+ }
766
+ },
767
+ "relative_action": {}
768
+ }
769
+ }
experiment_cfg/final_model_config.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_type": "Gr00tN1d7",
3
+ "model_dtype": "bfloat16",
4
+ "model_name": "nvidia/Cosmos-Reason2-2B",
5
+ "backbone_model_type": "qwen",
6
+ "model_revision": null,
7
+ "tune_top_llm_layers": 0,
8
+ "backbone_embedding_dim": 2048,
9
+ "tune_llm": false,
10
+ "tune_visual": false,
11
+ "select_layer": 16,
12
+ "reproject_vision": false,
13
+ "use_flash_attention": true,
14
+ "load_bf16": false,
15
+ "backbone_trainable_params_fp32": true,
16
+ "letter_box_transform": false,
17
+ "extra_augmentation_config": null,
18
+ "attn_supervision_layers": null,
19
+ "attn_supervision_weight": 0.0,
20
+ "attn_supervision_loss": "kl",
21
+ "attn_supervision_margin": 0.9,
22
+ "apply_sincos_state_encoding": false,
23
+ "use_percentiles": true,
24
+ "use_relative_action": false,
25
+ "max_state_dim": 132,
26
+ "max_action_dim": 132,
27
+ "action_horizon": 40,
28
+ "hidden_size": 1024,
29
+ "input_embedding_dim": 1536,
30
+ "state_history_length": 1,
31
+ "add_pos_embed": true,
32
+ "attn_dropout": 0.2,
33
+ "use_vlln": true,
34
+ "max_seq_len": 1024,
35
+ "use_alternate_vl_dit": true,
36
+ "attend_text_every_n_blocks": 2,
37
+ "diffusion_model_cfg": {
38
+ "attention_head_dim": 48,
39
+ "dropout": 0.2,
40
+ "final_dropout": true,
41
+ "interleave_self_attention": true,
42
+ "norm_type": "ada_norm",
43
+ "num_attention_heads": 32,
44
+ "num_layers": 32,
45
+ "output_dim": 1024,
46
+ "positional_embeddings": null
47
+ },
48
+ "num_inference_timesteps": 4,
49
+ "noise_beta_alpha": 1.5,
50
+ "noise_beta_beta": 1.0,
51
+ "noise_s": 0.999,
52
+ "num_timestep_buckets": 1000,
53
+ "ttrtc_enabled": false,
54
+ "ttrtc_max_delay_steps": 10,
55
+ "ttrtc_zero_delay_probability": 0.1,
56
+ "ttrtc_policy_hz": 15.0,
57
+ "tune_projector": true,
58
+ "tune_diffusion_model": true,
59
+ "tune_vlln": true,
60
+ "state_dropout_prob": 0.2,
61
+ "exclude_state": false,
62
+ "use_mean_std": false,
63
+ "max_num_embodiments": 32
64
+ }
experiment_cfg/final_processor_config.json ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbcaea5ee88f1e0f1465043920a2647c67e7de17d24adfd1c477742a6168edec
3
+ size 4986649584
model-00002-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c21ba9a5adae60f128f6d26a743155427bcd6b41490eed0eb5e26c4b0d7b94ac
3
+ size 4970792616
model-00003-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d263b62063a3fc64c314484241ae63c8713d78795b84f6b309a439ebba98667
3
+ size 2618758696
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f074cb1c989517e3ebc4c800d28b7f116a035f8b949f76e9951296a37eb1447e
3
+ size 3304120883
processor_config.json ADDED
@@ -0,0 +1,1129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "processor_class": "Gr00tN1d7Processor",
3
+ "processor_kwargs": {
4
+ "modality_configs": {
5
+ "real_g1_relative_eef_relative_joints": {
6
+ "video": {
7
+ "delta_indices": [
8
+ -20,
9
+ 0
10
+ ],
11
+ "modality_keys": [
12
+ "ego_view"
13
+ ],
14
+ "sin_cos_embedding_keys": null,
15
+ "mean_std_embedding_keys": null,
16
+ "action_configs": null
17
+ },
18
+ "state": {
19
+ "delta_indices": [
20
+ 0
21
+ ],
22
+ "modality_keys": [
23
+ "left_wrist_eef_9d",
24
+ "right_wrist_eef_9d",
25
+ "left_hand",
26
+ "right_hand",
27
+ "left_arm",
28
+ "right_arm",
29
+ "waist"
30
+ ],
31
+ "sin_cos_embedding_keys": null,
32
+ "mean_std_embedding_keys": null,
33
+ "action_configs": null
34
+ },
35
+ "action": {
36
+ "delta_indices": [
37
+ 0,
38
+ 1,
39
+ 2,
40
+ 3,
41
+ 4,
42
+ 5,
43
+ 6,
44
+ 7,
45
+ 8,
46
+ 9,
47
+ 10,
48
+ 11,
49
+ 12,
50
+ 13,
51
+ 14,
52
+ 15,
53
+ 16,
54
+ 17,
55
+ 18,
56
+ 19,
57
+ 20,
58
+ 21,
59
+ 22,
60
+ 23,
61
+ 24,
62
+ 25,
63
+ 26,
64
+ 27,
65
+ 28,
66
+ 29,
67
+ 30,
68
+ 31,
69
+ 32,
70
+ 33,
71
+ 34,
72
+ 35,
73
+ 36,
74
+ 37,
75
+ 38,
76
+ 39
77
+ ],
78
+ "modality_keys": [
79
+ "left_wrist_eef_9d",
80
+ "right_wrist_eef_9d",
81
+ "left_hand",
82
+ "right_hand",
83
+ "left_arm",
84
+ "right_arm",
85
+ "waist",
86
+ "base_height_command",
87
+ "navigate_command"
88
+ ],
89
+ "sin_cos_embedding_keys": null,
90
+ "mean_std_embedding_keys": null,
91
+ "action_configs": [
92
+ {
93
+ "rep": "RELATIVE",
94
+ "type": "EEF",
95
+ "format": "XYZ_ROT6D",
96
+ "state_key": "left_wrist_eef_9d"
97
+ },
98
+ {
99
+ "rep": "RELATIVE",
100
+ "type": "EEF",
101
+ "format": "XYZ_ROT6D",
102
+ "state_key": "right_wrist_eef_9d"
103
+ },
104
+ {
105
+ "rep": "ABSOLUTE",
106
+ "type": "NON_EEF",
107
+ "format": "DEFAULT",
108
+ "state_key": "left_hand"
109
+ },
110
+ {
111
+ "rep": "ABSOLUTE",
112
+ "type": "NON_EEF",
113
+ "format": "DEFAULT",
114
+ "state_key": "right_hand"
115
+ },
116
+ {
117
+ "rep": "RELATIVE",
118
+ "type": "NON_EEF",
119
+ "format": "DEFAULT",
120
+ "state_key": "left_arm"
121
+ },
122
+ {
123
+ "rep": "RELATIVE",
124
+ "type": "NON_EEF",
125
+ "format": "DEFAULT",
126
+ "state_key": "right_arm"
127
+ },
128
+ {
129
+ "rep": "ABSOLUTE",
130
+ "type": "NON_EEF",
131
+ "format": "DEFAULT",
132
+ "state_key": "waist"
133
+ },
134
+ {
135
+ "rep": "ABSOLUTE",
136
+ "type": "NON_EEF",
137
+ "format": "DEFAULT",
138
+ "state_key": "base_height_command"
139
+ },
140
+ {
141
+ "rep": "ABSOLUTE",
142
+ "type": "NON_EEF",
143
+ "format": "DEFAULT",
144
+ "state_key": "navigate_command"
145
+ }
146
+ ]
147
+ },
148
+ "language": {
149
+ "delta_indices": [
150
+ 0
151
+ ],
152
+ "modality_keys": [
153
+ "annotation.human.task_description"
154
+ ],
155
+ "sin_cos_embedding_keys": null,
156
+ "mean_std_embedding_keys": null,
157
+ "action_configs": null
158
+ }
159
+ },
160
+ "real_r1_pro_sharpa_relative_eef_mecka": {
161
+ "video": {
162
+ "delta_indices": [
163
+ -30,
164
+ 0
165
+ ],
166
+ "modality_keys": [
167
+ "ego_view_cropratio_res320x240_freq30"
168
+ ],
169
+ "sin_cos_embedding_keys": null,
170
+ "mean_std_embedding_keys": null,
171
+ "action_configs": null
172
+ },
173
+ "state": {
174
+ "delta_indices": [
175
+ 0
176
+ ],
177
+ "modality_keys": [
178
+ "left_wrist_eef",
179
+ "right_wrist_eef",
180
+ "left_hand_joints",
181
+ "right_hand_joints"
182
+ ],
183
+ "sin_cos_embedding_keys": null,
184
+ "mean_std_embedding_keys": null,
185
+ "action_configs": null
186
+ },
187
+ "action": {
188
+ "delta_indices": [
189
+ 0,
190
+ 1,
191
+ 2,
192
+ 3,
193
+ 4,
194
+ 5,
195
+ 6,
196
+ 7,
197
+ 8,
198
+ 9,
199
+ 10,
200
+ 11,
201
+ 12,
202
+ 13,
203
+ 14,
204
+ 15,
205
+ 16,
206
+ 17,
207
+ 18,
208
+ 19,
209
+ 20,
210
+ 21,
211
+ 22,
212
+ 23,
213
+ 24,
214
+ 25,
215
+ 26,
216
+ 27,
217
+ 28,
218
+ 29,
219
+ 30,
220
+ 31,
221
+ 32,
222
+ 33,
223
+ 34,
224
+ 35,
225
+ 36,
226
+ 37,
227
+ 38,
228
+ 39
229
+ ],
230
+ "modality_keys": [
231
+ "left_wrist_eef",
232
+ "right_wrist_eef",
233
+ "left_hand_joints",
234
+ "right_hand_joints"
235
+ ],
236
+ "sin_cos_embedding_keys": null,
237
+ "mean_std_embedding_keys": null,
238
+ "action_configs": [
239
+ {
240
+ "rep": "RELATIVE",
241
+ "type": "EEF",
242
+ "format": "XYZ_ROT6D",
243
+ "state_key": "left_wrist_eef"
244
+ },
245
+ {
246
+ "rep": "RELATIVE",
247
+ "type": "EEF",
248
+ "format": "XYZ_ROT6D",
249
+ "state_key": "right_wrist_eef"
250
+ },
251
+ {
252
+ "rep": "ABSOLUTE",
253
+ "type": "NON_EEF",
254
+ "format": "DEFAULT",
255
+ "state_key": "left_hand_joints"
256
+ },
257
+ {
258
+ "rep": "ABSOLUTE",
259
+ "type": "NON_EEF",
260
+ "format": "DEFAULT",
261
+ "state_key": "right_hand_joints"
262
+ }
263
+ ]
264
+ },
265
+ "language": {
266
+ "delta_indices": [
267
+ 0
268
+ ],
269
+ "modality_keys": [
270
+ "annotation.human.coarse_action"
271
+ ],
272
+ "sin_cos_embedding_keys": null,
273
+ "mean_std_embedding_keys": null,
274
+ "action_configs": null
275
+ }
276
+ },
277
+ "real_r1_pro_sharpa_relative_eef_human": {
278
+ "video": {
279
+ "delta_indices": [
280
+ -20,
281
+ 0
282
+ ],
283
+ "modality_keys": [
284
+ "ego_view_res320x240_freq20",
285
+ "left_wrist_view_res320x240_freq20",
286
+ "right_wrist_view_res320x240_freq20"
287
+ ],
288
+ "sin_cos_embedding_keys": null,
289
+ "mean_std_embedding_keys": null,
290
+ "action_configs": null
291
+ },
292
+ "state": {
293
+ "delta_indices": [
294
+ 0
295
+ ],
296
+ "modality_keys": [
297
+ "left_wrist_eef",
298
+ "right_wrist_eef",
299
+ "left_hand_joints",
300
+ "right_hand_joints"
301
+ ],
302
+ "sin_cos_embedding_keys": null,
303
+ "mean_std_embedding_keys": null,
304
+ "action_configs": null
305
+ },
306
+ "action": {
307
+ "delta_indices": [
308
+ 0,
309
+ 1,
310
+ 2,
311
+ 3,
312
+ 4,
313
+ 5,
314
+ 6,
315
+ 7,
316
+ 8,
317
+ 9,
318
+ 10,
319
+ 11,
320
+ 12,
321
+ 13,
322
+ 14,
323
+ 15,
324
+ 16,
325
+ 17,
326
+ 18,
327
+ 19,
328
+ 20,
329
+ 21,
330
+ 22,
331
+ 23,
332
+ 24,
333
+ 25,
334
+ 26,
335
+ 27,
336
+ 28,
337
+ 29,
338
+ 30,
339
+ 31,
340
+ 32,
341
+ 33,
342
+ 34,
343
+ 35,
344
+ 36,
345
+ 37,
346
+ 38,
347
+ 39
348
+ ],
349
+ "modality_keys": [
350
+ "left_wrist_eef",
351
+ "right_wrist_eef",
352
+ "left_hand_joints",
353
+ "right_hand_joints"
354
+ ],
355
+ "sin_cos_embedding_keys": null,
356
+ "mean_std_embedding_keys": null,
357
+ "action_configs": [
358
+ {
359
+ "rep": "RELATIVE",
360
+ "type": "EEF",
361
+ "format": "XYZ_ROT6D",
362
+ "state_key": "left_wrist_eef"
363
+ },
364
+ {
365
+ "rep": "RELATIVE",
366
+ "type": "EEF",
367
+ "format": "XYZ_ROT6D",
368
+ "state_key": "right_wrist_eef"
369
+ },
370
+ {
371
+ "rep": "ABSOLUTE",
372
+ "type": "NON_EEF",
373
+ "format": "DEFAULT",
374
+ "state_key": "left_hand_joints"
375
+ },
376
+ {
377
+ "rep": "ABSOLUTE",
378
+ "type": "NON_EEF",
379
+ "format": "DEFAULT",
380
+ "state_key": "right_hand_joints"
381
+ }
382
+ ]
383
+ },
384
+ "language": {
385
+ "delta_indices": [
386
+ 0
387
+ ],
388
+ "modality_keys": [
389
+ "annotation.human.coarse_action"
390
+ ],
391
+ "sin_cos_embedding_keys": null,
392
+ "mean_std_embedding_keys": null,
393
+ "action_configs": null
394
+ }
395
+ },
396
+ "real_r1_pro_sharpa_relative_eef": {
397
+ "video": {
398
+ "delta_indices": [
399
+ -20,
400
+ 0
401
+ ],
402
+ "modality_keys": [
403
+ "ego_view_res320x240_freq20",
404
+ "left_wrist_view_res320x240_freq20",
405
+ "right_wrist_view_res320x240_freq20"
406
+ ],
407
+ "sin_cos_embedding_keys": null,
408
+ "mean_std_embedding_keys": null,
409
+ "action_configs": null
410
+ },
411
+ "state": {
412
+ "delta_indices": [
413
+ 0
414
+ ],
415
+ "modality_keys": [
416
+ "left_wrist_eef",
417
+ "right_wrist_eef",
418
+ "left_hand_joints",
419
+ "right_hand_joints"
420
+ ],
421
+ "sin_cos_embedding_keys": null,
422
+ "mean_std_embedding_keys": null,
423
+ "action_configs": null
424
+ },
425
+ "action": {
426
+ "delta_indices": [
427
+ 0,
428
+ 1,
429
+ 2,
430
+ 3,
431
+ 4,
432
+ 5,
433
+ 6,
434
+ 7,
435
+ 8,
436
+ 9,
437
+ 10,
438
+ 11,
439
+ 12,
440
+ 13,
441
+ 14,
442
+ 15,
443
+ 16,
444
+ 17,
445
+ 18,
446
+ 19,
447
+ 20,
448
+ 21,
449
+ 22,
450
+ 23,
451
+ 24,
452
+ 25,
453
+ 26,
454
+ 27,
455
+ 28,
456
+ 29,
457
+ 30,
458
+ 31,
459
+ 32,
460
+ 33,
461
+ 34,
462
+ 35,
463
+ 36,
464
+ 37,
465
+ 38,
466
+ 39
467
+ ],
468
+ "modality_keys": [
469
+ "left_wrist_eef",
470
+ "right_wrist_eef",
471
+ "left_hand_joints",
472
+ "right_hand_joints"
473
+ ],
474
+ "sin_cos_embedding_keys": null,
475
+ "mean_std_embedding_keys": null,
476
+ "action_configs": [
477
+ {
478
+ "rep": "RELATIVE",
479
+ "type": "EEF",
480
+ "format": "XYZ_ROT6D",
481
+ "state_key": "left_wrist_eef"
482
+ },
483
+ {
484
+ "rep": "RELATIVE",
485
+ "type": "EEF",
486
+ "format": "XYZ_ROT6D",
487
+ "state_key": "right_wrist_eef"
488
+ },
489
+ {
490
+ "rep": "ABSOLUTE",
491
+ "type": "NON_EEF",
492
+ "format": "DEFAULT",
493
+ "state_key": "left_hand_joints"
494
+ },
495
+ {
496
+ "rep": "ABSOLUTE",
497
+ "type": "NON_EEF",
498
+ "format": "DEFAULT",
499
+ "state_key": "right_hand_joints"
500
+ }
501
+ ]
502
+ },
503
+ "language": {
504
+ "delta_indices": [
505
+ 0
506
+ ],
507
+ "modality_keys": [
508
+ "annotation.human.coarse_action"
509
+ ],
510
+ "sin_cos_embedding_keys": null,
511
+ "mean_std_embedding_keys": null,
512
+ "action_configs": null
513
+ }
514
+ },
515
+ "xdof_relative_eef_relative_joint": {
516
+ "video": {
517
+ "delta_indices": [
518
+ -30,
519
+ 0
520
+ ],
521
+ "modality_keys": [
522
+ "top_camera-images-rgb_320_240",
523
+ "left_camera-images-rgb_320_240",
524
+ "right_camera-images-rgb_320_240"
525
+ ],
526
+ "sin_cos_embedding_keys": null,
527
+ "mean_std_embedding_keys": null,
528
+ "action_configs": null
529
+ },
530
+ "state": {
531
+ "delta_indices": [
532
+ 0
533
+ ],
534
+ "modality_keys": [
535
+ "left_wrist_eef",
536
+ "right_wrist_eef",
537
+ "left_gripper_pos",
538
+ "right_gripper_pos",
539
+ "left_joint_pos",
540
+ "right_joint_pos"
541
+ ],
542
+ "sin_cos_embedding_keys": null,
543
+ "mean_std_embedding_keys": null,
544
+ "action_configs": null
545
+ },
546
+ "action": {
547
+ "delta_indices": [
548
+ 0,
549
+ 1,
550
+ 2,
551
+ 3,
552
+ 4,
553
+ 5,
554
+ 6,
555
+ 7,
556
+ 8,
557
+ 9,
558
+ 10,
559
+ 11,
560
+ 12,
561
+ 13,
562
+ 14,
563
+ 15,
564
+ 16,
565
+ 17,
566
+ 18,
567
+ 19,
568
+ 20,
569
+ 21,
570
+ 22,
571
+ 23,
572
+ 24,
573
+ 25,
574
+ 26,
575
+ 27,
576
+ 28,
577
+ 29,
578
+ 30,
579
+ 31,
580
+ 32,
581
+ 33,
582
+ 34,
583
+ 35,
584
+ 36,
585
+ 37,
586
+ 38,
587
+ 39
588
+ ],
589
+ "modality_keys": [
590
+ "left_wrist_eef",
591
+ "right_wrist_eef",
592
+ "left_gripper_pos",
593
+ "right_gripper_pos",
594
+ "left_joint_pos",
595
+ "right_joint_pos"
596
+ ],
597
+ "sin_cos_embedding_keys": null,
598
+ "mean_std_embedding_keys": null,
599
+ "action_configs": [
600
+ {
601
+ "rep": "RELATIVE",
602
+ "type": "EEF",
603
+ "format": "XYZ_ROT6D",
604
+ "state_key": "left_wrist_eef"
605
+ },
606
+ {
607
+ "rep": "RELATIVE",
608
+ "type": "EEF",
609
+ "format": "XYZ_ROT6D",
610
+ "state_key": "right_wrist_eef"
611
+ },
612
+ {
613
+ "rep": "ABSOLUTE",
614
+ "type": "NON_EEF",
615
+ "format": "DEFAULT",
616
+ "state_key": "left_gripper_pos"
617
+ },
618
+ {
619
+ "rep": "ABSOLUTE",
620
+ "type": "NON_EEF",
621
+ "format": "DEFAULT",
622
+ "state_key": "right_gripper_pos"
623
+ },
624
+ {
625
+ "rep": "RELATIVE",
626
+ "type": "NON_EEF",
627
+ "format": "DEFAULT",
628
+ "state_key": "left_joint_pos"
629
+ },
630
+ {
631
+ "rep": "RELATIVE",
632
+ "type": "NON_EEF",
633
+ "format": "DEFAULT",
634
+ "state_key": "right_joint_pos"
635
+ }
636
+ ]
637
+ },
638
+ "language": {
639
+ "delta_indices": [
640
+ 0
641
+ ],
642
+ "modality_keys": [
643
+ "annotation.task"
644
+ ],
645
+ "sin_cos_embedding_keys": null,
646
+ "mean_std_embedding_keys": null,
647
+ "action_configs": null
648
+ }
649
+ },
650
+ "real_r1_pro_sharpa_relative_eef_maxinsights": {
651
+ "video": {
652
+ "delta_indices": [
653
+ -30,
654
+ 0
655
+ ],
656
+ "modality_keys": [
657
+ "ego_view_cropratio_res320x240_freq30"
658
+ ],
659
+ "sin_cos_embedding_keys": null,
660
+ "mean_std_embedding_keys": null,
661
+ "action_configs": null
662
+ },
663
+ "state": {
664
+ "delta_indices": [
665
+ 0
666
+ ],
667
+ "modality_keys": [
668
+ "left_wrist_eef",
669
+ "right_wrist_eef",
670
+ "left_hand_joints",
671
+ "right_hand_joints"
672
+ ],
673
+ "sin_cos_embedding_keys": null,
674
+ "mean_std_embedding_keys": null,
675
+ "action_configs": null
676
+ },
677
+ "action": {
678
+ "delta_indices": [
679
+ 0,
680
+ 1,
681
+ 2,
682
+ 3,
683
+ 4,
684
+ 5,
685
+ 6,
686
+ 7,
687
+ 8,
688
+ 9,
689
+ 10,
690
+ 11,
691
+ 12,
692
+ 13,
693
+ 14,
694
+ 15,
695
+ 16,
696
+ 17,
697
+ 18,
698
+ 19,
699
+ 20,
700
+ 21,
701
+ 22,
702
+ 23,
703
+ 24,
704
+ 25,
705
+ 26,
706
+ 27,
707
+ 28,
708
+ 29,
709
+ 30,
710
+ 31,
711
+ 32,
712
+ 33,
713
+ 34,
714
+ 35,
715
+ 36,
716
+ 37,
717
+ 38,
718
+ 39
719
+ ],
720
+ "modality_keys": [
721
+ "left_wrist_eef",
722
+ "right_wrist_eef",
723
+ "left_hand_joints",
724
+ "right_hand_joints"
725
+ ],
726
+ "sin_cos_embedding_keys": null,
727
+ "mean_std_embedding_keys": null,
728
+ "action_configs": [
729
+ {
730
+ "rep": "RELATIVE",
731
+ "type": "EEF",
732
+ "format": "XYZ_ROT6D",
733
+ "state_key": "left_wrist_eef"
734
+ },
735
+ {
736
+ "rep": "RELATIVE",
737
+ "type": "EEF",
738
+ "format": "XYZ_ROT6D",
739
+ "state_key": "right_wrist_eef"
740
+ },
741
+ {
742
+ "rep": "ABSOLUTE",
743
+ "type": "NON_EEF",
744
+ "format": "DEFAULT",
745
+ "state_key": "left_hand_joints"
746
+ },
747
+ {
748
+ "rep": "ABSOLUTE",
749
+ "type": "NON_EEF",
750
+ "format": "DEFAULT",
751
+ "state_key": "right_hand_joints"
752
+ }
753
+ ]
754
+ },
755
+ "language": {
756
+ "delta_indices": [
757
+ 0
758
+ ],
759
+ "modality_keys": [
760
+ "annotation.human.coarse_action"
761
+ ],
762
+ "sin_cos_embedding_keys": null,
763
+ "mean_std_embedding_keys": null,
764
+ "action_configs": null
765
+ }
766
+ },
767
+ "xdof_relative_eef_relative_joint_subtask": {
768
+ "video": {
769
+ "delta_indices": [
770
+ -30,
771
+ 0
772
+ ],
773
+ "modality_keys": [
774
+ "top_camera-images-rgb_320_240",
775
+ "left_camera-images-rgb_320_240",
776
+ "right_camera-images-rgb_320_240"
777
+ ],
778
+ "sin_cos_embedding_keys": null,
779
+ "mean_std_embedding_keys": null,
780
+ "action_configs": null
781
+ },
782
+ "state": {
783
+ "delta_indices": [
784
+ 0
785
+ ],
786
+ "modality_keys": [
787
+ "left_wrist_eef",
788
+ "right_wrist_eef",
789
+ "left_gripper_pos",
790
+ "right_gripper_pos",
791
+ "left_joint_pos",
792
+ "right_joint_pos"
793
+ ],
794
+ "sin_cos_embedding_keys": null,
795
+ "mean_std_embedding_keys": null,
796
+ "action_configs": null
797
+ },
798
+ "action": {
799
+ "delta_indices": [
800
+ 0,
801
+ 1,
802
+ 2,
803
+ 3,
804
+ 4,
805
+ 5,
806
+ 6,
807
+ 7,
808
+ 8,
809
+ 9,
810
+ 10,
811
+ 11,
812
+ 12,
813
+ 13,
814
+ 14,
815
+ 15,
816
+ 16,
817
+ 17,
818
+ 18,
819
+ 19,
820
+ 20,
821
+ 21,
822
+ 22,
823
+ 23,
824
+ 24,
825
+ 25,
826
+ 26,
827
+ 27,
828
+ 28,
829
+ 29,
830
+ 30,
831
+ 31,
832
+ 32,
833
+ 33,
834
+ 34,
835
+ 35,
836
+ 36,
837
+ 37,
838
+ 38,
839
+ 39
840
+ ],
841
+ "modality_keys": [
842
+ "left_wrist_eef",
843
+ "right_wrist_eef",
844
+ "left_gripper_pos",
845
+ "right_gripper_pos",
846
+ "left_joint_pos",
847
+ "right_joint_pos"
848
+ ],
849
+ "sin_cos_embedding_keys": null,
850
+ "mean_std_embedding_keys": null,
851
+ "action_configs": [
852
+ {
853
+ "rep": "RELATIVE",
854
+ "type": "EEF",
855
+ "format": "XYZ_ROT6D",
856
+ "state_key": "left_wrist_eef"
857
+ },
858
+ {
859
+ "rep": "RELATIVE",
860
+ "type": "EEF",
861
+ "format": "XYZ_ROT6D",
862
+ "state_key": "right_wrist_eef"
863
+ },
864
+ {
865
+ "rep": "ABSOLUTE",
866
+ "type": "NON_EEF",
867
+ "format": "DEFAULT",
868
+ "state_key": "left_gripper_pos"
869
+ },
870
+ {
871
+ "rep": "ABSOLUTE",
872
+ "type": "NON_EEF",
873
+ "format": "DEFAULT",
874
+ "state_key": "right_gripper_pos"
875
+ },
876
+ {
877
+ "rep": "RELATIVE",
878
+ "type": "NON_EEF",
879
+ "format": "DEFAULT",
880
+ "state_key": "left_joint_pos"
881
+ },
882
+ {
883
+ "rep": "RELATIVE",
884
+ "type": "NON_EEF",
885
+ "format": "DEFAULT",
886
+ "state_key": "right_joint_pos"
887
+ }
888
+ ]
889
+ },
890
+ "language": {
891
+ "delta_indices": [
892
+ 0
893
+ ],
894
+ "modality_keys": [
895
+ "annotation.sub_task"
896
+ ],
897
+ "sin_cos_embedding_keys": null,
898
+ "mean_std_embedding_keys": null,
899
+ "action_configs": null
900
+ }
901
+ },
902
+ "oxe_droid_relative_eef_relative_joint": {
903
+ "video": {
904
+ "delta_indices": [
905
+ -15,
906
+ 0
907
+ ],
908
+ "modality_keys": [
909
+ "exterior_image_1_left",
910
+ "wrist_image_left"
911
+ ],
912
+ "sin_cos_embedding_keys": null,
913
+ "mean_std_embedding_keys": null,
914
+ "action_configs": null
915
+ },
916
+ "state": {
917
+ "delta_indices": [
918
+ 0
919
+ ],
920
+ "modality_keys": [
921
+ "eef_9d",
922
+ "gripper_position",
923
+ "joint_position"
924
+ ],
925
+ "sin_cos_embedding_keys": null,
926
+ "mean_std_embedding_keys": null,
927
+ "action_configs": null
928
+ },
929
+ "action": {
930
+ "delta_indices": [
931
+ 0,
932
+ 1,
933
+ 2,
934
+ 3,
935
+ 4,
936
+ 5,
937
+ 6,
938
+ 7,
939
+ 8,
940
+ 9,
941
+ 10,
942
+ 11,
943
+ 12,
944
+ 13,
945
+ 14,
946
+ 15,
947
+ 16,
948
+ 17,
949
+ 18,
950
+ 19,
951
+ 20,
952
+ 21,
953
+ 22,
954
+ 23,
955
+ 24,
956
+ 25,
957
+ 26,
958
+ 27,
959
+ 28,
960
+ 29,
961
+ 30,
962
+ 31,
963
+ 32,
964
+ 33,
965
+ 34,
966
+ 35,
967
+ 36,
968
+ 37,
969
+ 38,
970
+ 39
971
+ ],
972
+ "modality_keys": [
973
+ "eef_9d",
974
+ "gripper_position",
975
+ "joint_position"
976
+ ],
977
+ "sin_cos_embedding_keys": null,
978
+ "mean_std_embedding_keys": null,
979
+ "action_configs": [
980
+ {
981
+ "rep": "RELATIVE",
982
+ "type": "EEF",
983
+ "format": "XYZ_ROT6D",
984
+ "state_key": "eef_9d"
985
+ },
986
+ {
987
+ "rep": "ABSOLUTE",
988
+ "type": "NON_EEF",
989
+ "format": "DEFAULT",
990
+ "state_key": "gripper_position"
991
+ },
992
+ {
993
+ "rep": "RELATIVE",
994
+ "type": "NON_EEF",
995
+ "format": "DEFAULT",
996
+ "state_key": "joint_position"
997
+ }
998
+ ]
999
+ },
1000
+ "language": {
1001
+ "delta_indices": [
1002
+ 0
1003
+ ],
1004
+ "modality_keys": [
1005
+ "annotation.language.language_instruction"
1006
+ ],
1007
+ "sin_cos_embedding_keys": null,
1008
+ "mean_std_embedding_keys": null,
1009
+ "action_configs": null
1010
+ }
1011
+ },
1012
+ "new_embodiment": {
1013
+ "video": {
1014
+ "delta_indices": [
1015
+ 0
1016
+ ],
1017
+ "modality_keys": [
1018
+ "cam_left_head",
1019
+ "cam_left_wrist",
1020
+ "cam_right_wrist"
1021
+ ],
1022
+ "sin_cos_embedding_keys": null,
1023
+ "mean_std_embedding_keys": null,
1024
+ "action_configs": null
1025
+ },
1026
+ "state": {
1027
+ "delta_indices": [
1028
+ 0
1029
+ ],
1030
+ "modality_keys": [
1031
+ "arm_left",
1032
+ "arm_right",
1033
+ "hand_left",
1034
+ "hand_right"
1035
+ ],
1036
+ "sin_cos_embedding_keys": null,
1037
+ "mean_std_embedding_keys": null,
1038
+ "action_configs": null
1039
+ },
1040
+ "action": {
1041
+ "delta_indices": [
1042
+ 0,
1043
+ 1,
1044
+ 2,
1045
+ 3,
1046
+ 4,
1047
+ 5,
1048
+ 6,
1049
+ 7,
1050
+ 8,
1051
+ 9,
1052
+ 10,
1053
+ 11,
1054
+ 12,
1055
+ 13,
1056
+ 14,
1057
+ 15
1058
+ ],
1059
+ "modality_keys": [
1060
+ "arm_left",
1061
+ "arm_right",
1062
+ "hand_left",
1063
+ "hand_right"
1064
+ ],
1065
+ "sin_cos_embedding_keys": null,
1066
+ "mean_std_embedding_keys": null,
1067
+ "action_configs": [
1068
+ {
1069
+ "rep": "ABSOLUTE",
1070
+ "type": "NON_EEF",
1071
+ "format": "DEFAULT",
1072
+ "state_key": null
1073
+ },
1074
+ {
1075
+ "rep": "ABSOLUTE",
1076
+ "type": "NON_EEF",
1077
+ "format": "DEFAULT",
1078
+ "state_key": null
1079
+ },
1080
+ {
1081
+ "rep": "ABSOLUTE",
1082
+ "type": "NON_EEF",
1083
+ "format": "DEFAULT",
1084
+ "state_key": null
1085
+ },
1086
+ {
1087
+ "rep": "ABSOLUTE",
1088
+ "type": "NON_EEF",
1089
+ "format": "DEFAULT",
1090
+ "state_key": null
1091
+ }
1092
+ ]
1093
+ },
1094
+ "language": {
1095
+ "delta_indices": [
1096
+ 0
1097
+ ],
1098
+ "modality_keys": [
1099
+ "sub_task"
1100
+ ],
1101
+ "sin_cos_embedding_keys": null,
1102
+ "mean_std_embedding_keys": null,
1103
+ "action_configs": null
1104
+ }
1105
+ }
1106
+ },
1107
+ "image_crop_size": null,
1108
+ "image_target_size": null,
1109
+ "use_albumentations": true,
1110
+ "random_rotation_angle": null,
1111
+ "color_jitter_params": null,
1112
+ "shortest_image_edge": 256,
1113
+ "crop_fraction": 0.9,
1114
+ "letter_box_transform": true,
1115
+ "model_name": "nvidia/Cosmos-Reason2-2B",
1116
+ "model_type": "qwen",
1117
+ "formalize_language": true,
1118
+ "max_state_dim": 132,
1119
+ "max_action_dim": 132,
1120
+ "max_action_horizon": 40,
1121
+ "use_percentiles": true,
1122
+ "use_mean_std": false,
1123
+ "clip_outliers": true,
1124
+ "apply_sincos_state_encoding": false,
1125
+ "use_relative_action": true,
1126
+ "exclude_state": false,
1127
+ "state_dropout_prob": 0.2
1128
+ }
1129
+ }
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f438c30ddd3487a5672a44bcf7ecd404da010cb4ba699295ac1f48164f1517e
3
+ size 14645
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0994ada9df0673f832115c2e786787db3d753b809b380c5b150a0f4476ff9e9
3
+ size 1465
statistics.json ADDED
The diff for this file is too large to render. See raw diff
 
trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a15f86f492f43a1a1c9aea4bbbb1d9a8e318cebcab7b74c83331d4b9a6a921c
3
+ size 5969
wandb_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"project": "finetune-gr00t-n1d7", "run_id": "task000802_sh5_groot_n17_standard_100k_v1"}