solarian-diffusion 0.1.0__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (159) hide show
  1. diffusion_policy/__init__.py +0 -0
  2. diffusion_policy/codecs/__init__.py +0 -0
  3. diffusion_policy/codecs/imagecodecs_numcodecs.py +1408 -0
  4. diffusion_policy/common/__init__.py +0 -0
  5. diffusion_policy/common/checkpoint_util.py +64 -0
  6. diffusion_policy/common/json_logger.py +117 -0
  7. diffusion_policy/common/normalize_util.py +251 -0
  8. diffusion_policy/common/pose_repr_util.py +112 -0
  9. diffusion_policy/common/pytorch_util.py +82 -0
  10. diffusion_policy/common/q_critic_util.py +135 -0
  11. diffusion_policy/common/replay_buffer.py +672 -0
  12. diffusion_policy/common/sampler.py +204 -0
  13. diffusion_policy/config/legacy/task/blockpush_lowdim_seed.yaml +34 -0
  14. diffusion_policy/config/legacy/task/blockpush_lowdim_seed_abs.yaml +34 -0
  15. diffusion_policy/config/legacy/task/can_image.yaml +64 -0
  16. diffusion_policy/config/legacy/task/can_image_abs.yaml +64 -0
  17. diffusion_policy/config/legacy/task/can_lowdim.yaml +45 -0
  18. diffusion_policy/config/legacy/task/can_lowdim_abs.yaml +46 -0
  19. diffusion_policy/config/legacy/task/kitchen_lowdim.yaml +33 -0
  20. diffusion_policy/config/legacy/task/kitchen_lowdim_abs.yaml +38 -0
  21. diffusion_policy/config/legacy/task/lift_abs.yaml +72 -0
  22. diffusion_policy/config/legacy/task/lift_image.yaml +64 -0
  23. diffusion_policy/config/legacy/task/lift_image_abs.yaml +63 -0
  24. diffusion_policy/config/legacy/task/lift_lowdim.yaml +46 -0
  25. diffusion_policy/config/legacy/task/lift_lowdim_abs.yaml +47 -0
  26. diffusion_policy/config/legacy/task/pusht_image.yaml +40 -0
  27. diffusion_policy/config/legacy/task/pusht_lowdim.yaml +34 -0
  28. diffusion_policy/config/legacy/task/real_pusht_image.yaml +47 -0
  29. diffusion_policy/config/legacy/task/square_abs.yaml +72 -0
  30. diffusion_policy/config/legacy/task/square_image.yaml +64 -0
  31. diffusion_policy/config/legacy/task/square_image_abs.yaml +64 -0
  32. diffusion_policy/config/legacy/task/square_lowdim.yaml +46 -0
  33. diffusion_policy/config/legacy/task/square_lowdim_abs.yaml +47 -0
  34. diffusion_policy/config/legacy/task/tool_hang_abs.yaml +71 -0
  35. diffusion_policy/config/legacy/task/tool_hang_image.yaml +63 -0
  36. diffusion_policy/config/legacy/task/tool_hang_image_abs.yaml +63 -0
  37. diffusion_policy/config/legacy/task/tool_hang_lowdim.yaml +45 -0
  38. diffusion_policy/config/legacy/task/tool_hang_lowdim_abs.yaml +46 -0
  39. diffusion_policy/config/legacy/task/transport_image.yaml +75 -0
  40. diffusion_policy/config/legacy/task/transport_image_abs.yaml +75 -0
  41. diffusion_policy/config/legacy/task/transport_lowdim.yaml +49 -0
  42. diffusion_policy/config/legacy/task/transport_lowdim_abs.yaml +49 -0
  43. diffusion_policy/config/legacy/task/umi.yaml +45 -0
  44. diffusion_policy/config/legacy/task/umi_image.yaml +36 -0
  45. diffusion_policy/config/legacy/task/umi_teleop_image.yaml +41 -0
  46. diffusion_policy/config/legacy/train_bet_lowdim_workspace.yaml +132 -0
  47. diffusion_policy/config/legacy/train_diffusion_transformer_hybrid_workspace.yaml +141 -0
  48. diffusion_policy/config/legacy/train_diffusion_transformer_lowdim_kitchen_workspace.yaml +146 -0
  49. diffusion_policy/config/legacy/train_diffusion_transformer_lowdim_pusht_workspace.yaml +146 -0
  50. diffusion_policy/config/legacy/train_diffusion_transformer_lowdim_workspace.yaml +146 -0
  51. diffusion_policy/config/legacy/train_diffusion_transformer_real_hybrid_workspace.yaml +144 -0
  52. diffusion_policy/config/legacy/train_diffusion_transformer_umi_teleop_hybrid_workspace.yaml +156 -0
  53. diffusion_policy/config/legacy/train_diffusion_unet_ddim_hybrid_workspace.yaml +140 -0
  54. diffusion_policy/config/legacy/train_diffusion_unet_ddim_lowdim_workspace.yaml +146 -0
  55. diffusion_policy/config/legacy/train_diffusion_unet_hybrid_workspace.yaml +137 -0
  56. diffusion_policy/config/legacy/train_diffusion_unet_image_pretrained_workspace.yaml +149 -0
  57. diffusion_policy/config/legacy/train_diffusion_unet_image_workspace.yaml +150 -0
  58. diffusion_policy/config/legacy/train_diffusion_unet_lowdim_workspace.yaml +143 -0
  59. diffusion_policy/config/legacy/train_diffusion_unet_real_hybrid_workspace.yaml +141 -0
  60. diffusion_policy/config/legacy/train_diffusion_unet_real_image_workspace.yaml +152 -0
  61. diffusion_policy/config/legacy/train_diffusion_unet_real_pretrained_workspace.yaml +152 -0
  62. diffusion_policy/config/legacy/train_diffusion_unet_timm_workspace.yaml +145 -0
  63. diffusion_policy/config/legacy/train_diffusion_unet_umi_hybrid_workspace.yaml +141 -0
  64. diffusion_policy/config/legacy/train_diffusion_unet_umi_teleop_hybrid_workspace.yaml +153 -0
  65. diffusion_policy/config/legacy/train_diffusion_unet_umi_timm_workspace.yaml +145 -0
  66. diffusion_policy/config/legacy/train_diffusion_unet_umi_workspace.yaml +150 -0
  67. diffusion_policy/config/legacy/train_diffusion_unet_video_workspace.yaml +128 -0
  68. diffusion_policy/config/legacy/train_ibc_dfo_hybrid_workspace.yaml +111 -0
  69. diffusion_policy/config/legacy/train_ibc_dfo_lowdim_workspace.yaml +109 -0
  70. diffusion_policy/config/legacy/train_ibc_dfo_real_hybrid_workspace.yaml +111 -0
  71. diffusion_policy/config/legacy/train_robomimic_image_workspace.yaml +92 -0
  72. diffusion_policy/config/legacy/train_robomimic_lowdim_workspace.yaml +92 -0
  73. diffusion_policy/config/legacy/train_robomimic_real_image_workspace.yaml +92 -0
  74. diffusion_policy/config/rl_annotate.yaml +265 -0
  75. diffusion_policy/config/rl_annotate_per_frame.yaml +162 -0
  76. diffusion_policy/config/rl_annotate_per_frame_v4.yaml +162 -0
  77. diffusion_policy/config/rl_annotate_v2.yaml +141 -0
  78. diffusion_policy/config/rl_annotate_v3.yaml +165 -0
  79. diffusion_policy/config/task/lift.yaml +89 -0
  80. diffusion_policy/config/task/square.yaml +92 -0
  81. diffusion_policy/config/task/tool_hang.yaml +93 -0
  82. diffusion_policy/config/task/tool_hang_abs.yaml +71 -0
  83. diffusion_policy/config/task/umi.yaml +105 -0
  84. diffusion_policy/config/task/umi_bimanual.yaml +135 -0
  85. diffusion_policy/config/task/umi_cup_wild.yaml +106 -0
  86. diffusion_policy/config/task/umi_image.yaml +36 -0
  87. diffusion_policy/config/task/umi_pick_lego.yaml +106 -0
  88. diffusion_policy/config/task/umi_pick_single_lego_175.yaml +106 -0
  89. diffusion_policy/config/task/umi_pick_single_lego_bbox.yaml +135 -0
  90. diffusion_policy/config/task/umi_teleop_image.yaml +41 -0
  91. diffusion_policy/config/train_diffusion_transformer_umi_bimanual_workspace.yaml +185 -0
  92. diffusion_policy/config/train_diffusion_transformer_umi_workspace.yaml +185 -0
  93. diffusion_policy/config/train_diffusion_unet_image_workspace.yaml +152 -0
  94. diffusion_policy/config/train_diffusion_unet_timm_robomimic_workspace.yaml +152 -0
  95. diffusion_policy/config/train_diffusion_unet_timm_umi_cup_wild.yaml +260 -0
  96. diffusion_policy/config/train_diffusion_unet_timm_umi_pick_lego_pile_bbox.yaml +265 -0
  97. diffusion_policy/config/train_diffusion_unet_timm_umi_pick_lego_pile_bbox_awr.yaml +324 -0
  98. diffusion_policy/config/train_diffusion_unet_timm_umi_pick_lego_pile_bbox_parl.yaml +399 -0
  99. diffusion_policy/config/train_diffusion_unet_timm_umi_pick_single_lego.yaml +258 -0
  100. diffusion_policy/config/train_diffusion_unet_timm_umi_pick_single_lego_125.yaml +260 -0
  101. diffusion_policy/config/train_diffusion_unet_timm_umi_pick_single_lego_175.yaml +260 -0
  102. diffusion_policy/config/train_diffusion_unet_timm_umi_pick_single_lego_bbox.yaml +265 -0
  103. diffusion_policy/config/train_diffusion_unet_timm_umi_pick_two_legos_bbox.yaml +265 -0
  104. diffusion_policy/config/train_diffusion_unet_timm_umi_workspace.yaml +224 -0
  105. diffusion_policy/config/train_diffusion_unet_timm_umi_workspace_lego.yaml +213 -0
  106. diffusion_policy/config/train_diffusion_unet_timm_umi_workspace_lego_resume.yaml +213 -0
  107. diffusion_policy/config/train_diffusion_unet_timm_umi_workspace_pick_lego.yaml +254 -0
  108. diffusion_policy/config/train_diffusion_unet_timm_umi_workspace_red_cup.yaml +219 -0
  109. diffusion_policy/config/train_diffusion_unet_umi_bimanual_workspace.yaml +203 -0
  110. diffusion_policy/config/train_iql_umi.yaml +186 -0
  111. diffusion_policy/config/train_iql_umi_v2.yaml +212 -0
  112. diffusion_policy/config/train_iql_umi_v3.yaml +212 -0
  113. diffusion_policy/dataset/__init__.py +0 -0
  114. diffusion_policy/dataset/base_dataset.py +76 -0
  115. diffusion_policy/dataset/umi_dataset.py +652 -0
  116. diffusion_policy/env_runner/__init__.py +0 -0
  117. diffusion_policy/env_runner/base_image_runner.py +9 -0
  118. diffusion_policy/env_runner/real_pusht_image_runner.py +10 -0
  119. diffusion_policy/model/__init__.py +0 -0
  120. diffusion_policy/model/common/__init__.py +0 -0
  121. diffusion_policy/model/common/dict_of_tensor_mixin.py +38 -0
  122. diffusion_policy/model/common/lr_scheduler.py +46 -0
  123. diffusion_policy/model/common/module_attr_mixin.py +14 -0
  124. diffusion_policy/model/common/normalizer.py +353 -0
  125. diffusion_policy/model/common/tensor_util.py +960 -0
  126. diffusion_policy/model/diffusion/__init__.py +0 -0
  127. diffusion_policy/model/diffusion/conditional_unet1d.py +238 -0
  128. diffusion_policy/model/diffusion/conv1d_components.py +46 -0
  129. diffusion_policy/model/diffusion/ema_model.py +88 -0
  130. diffusion_policy/model/diffusion/mask_generator.py +221 -0
  131. diffusion_policy/model/diffusion/positional_embedding.py +17 -0
  132. diffusion_policy/model/vision/__init__.py +0 -0
  133. diffusion_policy/model/vision/crop_randomizer.py +288 -0
  134. diffusion_policy/model/vision/multi_image_obs_encoder.py +195 -0
  135. diffusion_policy/model/vision/timm_obs_encoder.py +525 -0
  136. diffusion_policy/policy/__init__.py +0 -0
  137. diffusion_policy/policy/base_image_policy.py +27 -0
  138. diffusion_policy/policy/diffusion_unet_image_policy.py +264 -0
  139. diffusion_policy/policy/diffusion_unet_timm_policy.py +607 -0
  140. diffusion_policy/policy/parl_action_optimizer.py +675 -0
  141. diffusion_policy/workspace/__init__.py +0 -0
  142. diffusion_policy/workspace/base_workspace.py +146 -0
  143. diffusion_policy/workspace/train_diffusion_unet_image_workspace.py +1585 -0
  144. solarian_diffusion/__init__.py +32 -0
  145. solarian_diffusion/config.py +31 -0
  146. solarian_diffusion/policy_server.py +316 -0
  147. solarian_diffusion/rtc.py +272 -0
  148. solarian_diffusion/service.py +16 -0
  149. solarian_diffusion/trainer.py +213 -0
  150. solarian_diffusion-0.1.0.dist-info/METADATA +126 -0
  151. solarian_diffusion-0.1.0.dist-info/RECORD +159 -0
  152. solarian_diffusion-0.1.0.dist-info/WHEEL +5 -0
  153. solarian_diffusion-0.1.0.dist-info/entry_points.txt +8 -0
  154. solarian_diffusion-0.1.0.dist-info/licenses/LICENSE +201 -0
  155. solarian_diffusion-0.1.0.dist-info/licenses/NOTICE +45 -0
  156. solarian_diffusion-0.1.0.dist-info/top_level.txt +3 -0
  157. umi/__init__.py +0 -0
  158. umi/common/__init__.py +0 -0
  159. umi/common/pose_util.py +115 -0
@@ -0,0 +1,152 @@
1
+ defaults:
2
+ - _self_
3
+ - task: real_pusht_image
4
+
5
+ name: train_diffusion_unet_image
6
+ _target_: diffusion_policy.workspace.train_diffusion_unet_image_workspace.TrainDiffusionUnetImageWorkspace
7
+
8
+ task_name: ${task.name}
9
+ shape_meta: ${task.shape_meta}
10
+ exp_name: "default"
11
+
12
+ horizon: 16
13
+ n_obs_steps: 2
14
+ n_action_steps: 8
15
+ n_latency_steps: 0
16
+ dataset_obs_steps: ${n_obs_steps}
17
+ past_action_visible: False
18
+ keypoint_visible_rate: 1.0
19
+ obs_as_global_cond: True
20
+
21
+ policy:
22
+ _target_: diffusion_policy.policy.diffusion_unet_image_policy.DiffusionUnetImagePolicy
23
+
24
+ shape_meta: ${shape_meta}
25
+
26
+ noise_scheduler:
27
+ _target_: diffusers.schedulers.scheduling_ddim.DDIMScheduler
28
+ num_train_timesteps: 100
29
+ beta_start: 0.0001
30
+ beta_end: 0.02
31
+ # beta_schedule is important
32
+ # this is the best we found
33
+ beta_schedule: squaredcos_cap_v2
34
+ clip_sample: True
35
+ set_alpha_to_one: True
36
+ steps_offset: 0
37
+ prediction_type: epsilon # or sample
38
+
39
+ obs_encoder:
40
+ _target_: diffusion_policy.model.vision.multi_image_obs_encoder.MultiImageObsEncoder
41
+ shape_meta: ${shape_meta}
42
+ rgb_model:
43
+ _target_: diffusion_policy.model.vision.model_getter.get_resnet
44
+ name: resnet18
45
+ weights: null
46
+ resize_shape: [240, 320]
47
+ crop_shape: [216, 288] # ch, cw 240x320 90%
48
+ random_crop: True
49
+ use_group_norm: True
50
+ share_rgb_model: False
51
+ imagenet_norm: True
52
+
53
+ horizon: ${horizon}
54
+ n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
55
+ n_obs_steps: ${n_obs_steps}
56
+ num_inference_steps: 100
57
+ obs_as_global_cond: ${obs_as_global_cond}
58
+ # crop_shape: null
59
+ diffusion_step_embed_dim: 128
60
+ down_dims: [512, 1024, 2048]
61
+ kernel_size: 5
62
+ n_groups: 8
63
+ cond_predict_scale: True
64
+
65
+ # scheduler.step params
66
+ # predict_epsilon: True
67
+
68
+ ema:
69
+ _target_: diffusion_policy.model.diffusion.ema_model.EMAModel
70
+ update_after_step: 0
71
+ inv_gamma: 1.0
72
+ power: 0.75
73
+ min_value: 0.0
74
+ max_value: 0.9999
75
+
76
+ dataloader:
77
+ batch_size: 64
78
+ num_workers: 8
79
+ shuffle: True
80
+ pin_memory: True
81
+ persistent_workers: True
82
+
83
+ val_dataloader:
84
+ batch_size: 64
85
+ num_workers: 8
86
+ shuffle: False
87
+ pin_memory: True
88
+ persistent_workers: True
89
+
90
+ optimizer:
91
+ _target_: torch.optim.AdamW
92
+ lr: 1.0e-4
93
+ betas: [0.95, 0.999]
94
+ eps: 1.0e-8
95
+ weight_decay: 1.0e-6
96
+
97
+ training:
98
+ device: "cuda:0"
99
+ seed: 42
100
+ debug: False
101
+ resume: True
102
+ # optimization
103
+ lr_scheduler: cosine
104
+ lr_warmup_steps: 500
105
+ num_epochs: 600
106
+ gradient_accumulate_every: 1
107
+ # EMA destroys performance when used with BatchNorm
108
+ # replace BatchNorm with GroupNorm.
109
+ use_ema: True
110
+ freeze_encoder: False
111
+ # training loop control
112
+ # in epochs
113
+ rollout_every: 50
114
+ checkpoint_every: 50
115
+ val_every: 1
116
+ sample_every: 5
117
+ # steps per epoch
118
+ max_train_steps: null
119
+ max_val_steps: null
120
+ # misc
121
+ tqdm_interval_sec: 1.0
122
+
123
+ logging:
124
+ project: diffusion_policy_debug
125
+ resume: True
126
+ mode: online
127
+ name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
128
+ tags: ["${name}", "${task_name}", "${exp_name}"]
129
+ id: null
130
+ group: null
131
+
132
+ checkpoint:
133
+ topk:
134
+ monitor_key: train_loss
135
+ mode: min
136
+ k: 5
137
+ format_str: 'epoch={epoch:04d}-train_loss={train_loss:.3f}.ckpt'
138
+ save_last_ckpt: True
139
+ save_last_snapshot: False
140
+
141
+ multi_run:
142
+ run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
143
+ wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
144
+
145
+ hydra:
146
+ job:
147
+ override_dirname: ${name}
148
+ run:
149
+ dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
150
+ sweep:
151
+ dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
152
+ subdir: ${hydra.job.num}
@@ -0,0 +1,152 @@
1
+ defaults:
2
+ - _self_
3
+ - task: real_pusht_image
4
+
5
+ name: train_diffusion_unet_image
6
+ _target_: diffusion_policy.workspace.train_diffusion_unet_image_workspace.TrainDiffusionUnetImageWorkspace
7
+
8
+ task_name: ${task.name}
9
+ shape_meta: ${task.shape_meta}
10
+ exp_name: "default"
11
+
12
+ horizon: 16
13
+ n_obs_steps: 2
14
+ n_action_steps: 8
15
+ n_latency_steps: 0
16
+ dataset_obs_steps: ${n_obs_steps}
17
+ past_action_visible: False
18
+ keypoint_visible_rate: 1.0
19
+ obs_as_global_cond: True
20
+
21
+ policy:
22
+ _target_: diffusion_policy.policy.diffusion_unet_image_policy.DiffusionUnetImagePolicy
23
+
24
+ shape_meta: ${shape_meta}
25
+
26
+ noise_scheduler:
27
+ _target_: diffusers.schedulers.scheduling_ddim.DDIMScheduler
28
+ num_train_timesteps: 100
29
+ beta_start: 0.0001
30
+ beta_end: 0.02
31
+ # beta_schedule is important
32
+ # this is the best we found
33
+ beta_schedule: squaredcos_cap_v2
34
+ clip_sample: True
35
+ set_alpha_to_one: True
36
+ steps_offset: 0
37
+ prediction_type: epsilon # or sample
38
+
39
+ obs_encoder:
40
+ _target_: diffusion_policy.model.vision.multi_image_obs_encoder.MultiImageObsEncoder
41
+ shape_meta: ${shape_meta}
42
+ rgb_model:
43
+ _target_: diffusion_policy.model.vision.model_getter.get_resnet
44
+ name: resnet18
45
+ weights: IMAGENET1K_V1 # or r3m
46
+ resize_shape: [224,224]
47
+ crop_shape: null
48
+ random_crop: False
49
+ use_group_norm: False
50
+ share_rgb_model: True
51
+ imagenet_norm: True
52
+
53
+ horizon: ${horizon}
54
+ n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
55
+ n_obs_steps: ${n_obs_steps}
56
+ num_inference_steps: 100
57
+ obs_as_global_cond: ${obs_as_global_cond}
58
+ # crop_shape: null
59
+ diffusion_step_embed_dim: 128
60
+ down_dims: [512, 1024, 2048]
61
+ kernel_size: 5
62
+ n_groups: 8
63
+ cond_predict_scale: True
64
+
65
+ # scheduler.step params
66
+ # predict_epsilon: True
67
+
68
+ ema:
69
+ _target_: diffusion_policy.model.diffusion.ema_model.EMAModel
70
+ update_after_step: 0
71
+ inv_gamma: 1.0
72
+ power: 0.75
73
+ min_value: 0.0
74
+ max_value: 0.9999
75
+
76
+ dataloader:
77
+ batch_size: 64
78
+ num_workers: 8
79
+ shuffle: True
80
+ pin_memory: True
81
+ persistent_workers: True
82
+
83
+ val_dataloader:
84
+ batch_size: 64
85
+ num_workers: 8
86
+ shuffle: False
87
+ pin_memory: True
88
+ persistent_workers: True
89
+
90
+ optimizer:
91
+ _target_: torch.optim.AdamW
92
+ lr: 1.0e-4
93
+ betas: [0.95, 0.999]
94
+ eps: 1.0e-8
95
+ weight_decay: 1.0e-6
96
+
97
+ training:
98
+ device: "cuda:0"
99
+ seed: 42
100
+ debug: False
101
+ resume: True
102
+ # optimization
103
+ lr_scheduler: cosine
104
+ lr_warmup_steps: 500
105
+ num_epochs: 600
106
+ gradient_accumulate_every: 1
107
+ # EMA destroys performance when used with BatchNorm
108
+ # replace BatchNorm with GroupNorm.
109
+ use_ema: True
110
+ freeze_encoder: True
111
+ # training loop control
112
+ # in epochs
113
+ rollout_every: 50
114
+ checkpoint_every: 50
115
+ val_every: 1
116
+ sample_every: 5
117
+ # steps per epoch
118
+ max_train_steps: null
119
+ max_val_steps: null
120
+ # misc
121
+ tqdm_interval_sec: 1.0
122
+
123
+ logging:
124
+ project: diffusion_policy_debug
125
+ resume: True
126
+ mode: online
127
+ name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
128
+ tags: ["${name}", "${task_name}", "${exp_name}"]
129
+ id: null
130
+ group: null
131
+
132
+ checkpoint:
133
+ topk:
134
+ monitor_key: train_loss
135
+ mode: min
136
+ k: 5
137
+ format_str: 'epoch={epoch:04d}-train_loss={train_loss:.3f}.ckpt'
138
+ save_last_ckpt: True
139
+ save_last_snapshot: False
140
+
141
+ multi_run:
142
+ run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
143
+ wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
144
+
145
+ hydra:
146
+ job:
147
+ override_dirname: ${name}
148
+ run:
149
+ dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
150
+ sweep:
151
+ dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
152
+ subdir: ${hydra.job.num}
@@ -0,0 +1,145 @@
1
+ defaults:
2
+ - _self_
3
+ - task: lift_image_abs
4
+
5
+ name: train_diffusion_unet_timm
6
+ _target_: diffusion_policy.workspace.train_diffusion_unet_image_workspace.TrainDiffusionUnetImageWorkspace
7
+
8
+ task_name: ${task.name}
9
+ shape_meta: ${task.shape_meta}
10
+ exp_name: "default"
11
+
12
+ policy:
13
+ _target_: diffusion_policy.policy.diffusion_unet_timm_policy.DiffusionUnetTimmPolicy
14
+
15
+ shape_meta: ${shape_meta}
16
+
17
+ noise_scheduler:
18
+ _target_: diffusers.schedulers.scheduling_ddpm.DDPMScheduler
19
+ num_train_timesteps: 100
20
+ beta_start: 0.0001
21
+ beta_end: 0.02
22
+ beta_schedule: squaredcos_cap_v2
23
+ variance_type: fixed_small # Yilun's paper uses fixed_small_log instead, but easy to cause Nan
24
+ clip_sample: True # required when predict_epsilon=False
25
+ prediction_type: epsilon # or sample
26
+
27
+ obs_encoder:
28
+ _target_: diffusion_policy.model.vision.timm_obs_encoder.TimmObsEncoder
29
+ shape_meta: ${shape_meta}
30
+ model_name: 'resnet18'
31
+ # model_name: 'vit_base_patch16_224'
32
+ pretrained: False
33
+ global_pool: 'avg'
34
+ transforms:
35
+ - _target_: torchvision.transforms.Resize
36
+ size: 224
37
+ antialias: True
38
+ - _target_: torchvision.transforms.RandomRotation
39
+ degrees:
40
+ - -5.0
41
+ - 5.0
42
+ expand: false
43
+ - _target_: torchvision.transforms.ColorJitter
44
+ brightness: 0.2
45
+ contrast: 0.5
46
+ saturation: 0.2
47
+ hue: 0.2
48
+
49
+ use_group_norm: True
50
+ share_rgb_model: False
51
+ imagenet_norm: True
52
+
53
+ num_inference_steps: 100
54
+ obs_as_global_cond: True
55
+ diffusion_step_embed_dim: 128
56
+ down_dims: [512, 1024, 2048]
57
+ kernel_size: 5
58
+ n_groups: 8
59
+ cond_predict_scale: True
60
+
61
+ ema:
62
+ _target_: diffusion_policy.model.diffusion.ema_model.EMAModel
63
+ update_after_step: 0
64
+ inv_gamma: 1.0
65
+ power: 0.75
66
+ min_value: 0.0
67
+ max_value: 0.9999
68
+
69
+ dataloader:
70
+ batch_size: 64
71
+ num_workers: 4
72
+ shuffle: True
73
+ pin_memory: True
74
+ persistent_workers: False
75
+
76
+ val_dataloader:
77
+ batch_size: 64
78
+ num_workers: 4
79
+ shuffle: False
80
+ pin_memory: True
81
+ persistent_workers: False
82
+
83
+ optimizer:
84
+ _target_: torch.optim.AdamW
85
+ lr: 1.0e-4
86
+ betas: [0.95, 0.999]
87
+ eps: 1.0e-8
88
+ weight_decay: 1.0e-6
89
+
90
+ training:
91
+ device: "cuda:0"
92
+ seed: 42
93
+ debug: False
94
+ resume: True
95
+ # optimization
96
+ lr_scheduler: cosine
97
+ lr_warmup_steps: 500
98
+ num_epochs: 8000
99
+ gradient_accumulate_every: 1
100
+ # EMA destroys performance when used with BatchNorm
101
+ # replace BatchNorm with GroupNorm.
102
+ use_ema: True
103
+ freeze_encoder: False
104
+ # training loop control
105
+ # in epochs
106
+ rollout_every: 50
107
+ checkpoint_every: 50
108
+ val_every: 1
109
+ sample_every: 5
110
+ # steps per epoch
111
+ max_train_steps: null
112
+ max_val_steps: null
113
+ # misc
114
+ tqdm_interval_sec: 1.0
115
+
116
+ logging:
117
+ project: diffusion_policy_debug
118
+ resume: True
119
+ mode: online
120
+ name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
121
+ tags: ["${name}", "${task_name}", "${exp_name}"]
122
+ id: null
123
+ group: null
124
+
125
+ checkpoint:
126
+ topk:
127
+ monitor_key: test_mean_score
128
+ mode: max
129
+ k: 5
130
+ format_str: 'epoch={epoch:04d}-test_mean_score={test_mean_score:.3f}.ckpt'
131
+ save_last_ckpt: True
132
+ save_last_snapshot: False
133
+
134
+ multi_run:
135
+ run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
136
+ wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
137
+
138
+ hydra:
139
+ job:
140
+ override_dirname: ${name}
141
+ run:
142
+ dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
143
+ sweep:
144
+ dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
145
+ subdir: ${hydra.job.num}
@@ -0,0 +1,141 @@
1
+ defaults:
2
+ - _self_
3
+ - task: umi_image
4
+
5
+ name: train_diffusion_unet_hybrid
6
+ _target_: diffusion_policy.workspace.train_diffusion_unet_hybrid_workspace.TrainDiffusionUnetHybridWorkspace
7
+
8
+ task_name: ${task.name}
9
+ shape_meta: ${task.shape_meta}
10
+ exp_name: "default"
11
+
12
+ horizon: 32
13
+ n_obs_steps: 2
14
+ n_action_steps: 30
15
+ n_latency_steps: 0
16
+ dataset_obs_steps: ${n_obs_steps}
17
+ past_action_visible: False
18
+ keypoint_visible_rate: 1.0
19
+ obs_as_global_cond: True
20
+
21
+ policy:
22
+ _target_: diffusion_policy.policy.diffusion_unet_hybrid_image_policy.DiffusionUnetHybridImagePolicy
23
+
24
+ shape_meta: ${shape_meta}
25
+
26
+ noise_scheduler:
27
+ _target_: diffusers.schedulers.scheduling_ddim.DDIMScheduler
28
+ num_train_timesteps: 100
29
+ beta_start: 0.0001
30
+ beta_end: 0.02
31
+ # beta_schedule is important
32
+ # this is the best we found
33
+ beta_schedule: squaredcos_cap_v2
34
+ clip_sample: True
35
+ set_alpha_to_one: True
36
+ steps_offset: 0
37
+ prediction_type: epsilon # or sample
38
+
39
+ horizon: ${horizon}
40
+ n_action_steps: ${eval:'${n_action_steps}+${n_latency_steps}'}
41
+ n_obs_steps: ${n_obs_steps}
42
+ num_inference_steps: 8
43
+ obs_as_global_cond: ${obs_as_global_cond}
44
+ # crop_shape: [76, 76] # 84x84 90%
45
+ crop_shape: [216, 216] # ch, cw 320x240 90%
46
+ # crop_shape: null
47
+ diffusion_step_embed_dim: 128
48
+ down_dims: [256,512,1024]
49
+ kernel_size: 5
50
+ n_groups: 8
51
+ cond_predict_scale: True
52
+ obs_encoder_group_norm: True
53
+ eval_fixed_crop: True
54
+
55
+ # scheduler.step params
56
+ # predict_epsilon: True
57
+
58
+ ema:
59
+ _target_: diffusion_policy.model.diffusion.ema_model.EMAModel
60
+ update_after_step: 0
61
+ inv_gamma: 1.0
62
+ power: 0.75
63
+ min_value: 0.0
64
+ max_value: 0.9999
65
+
66
+ dataloader:
67
+ batch_size: 256
68
+ num_workers: 16
69
+ shuffle: True
70
+ pin_memory: True
71
+ persistent_workers: True
72
+
73
+ val_dataloader:
74
+ batch_size: 256
75
+ num_workers: 16
76
+ shuffle: False
77
+ pin_memory: True
78
+ persistent_workers: True
79
+
80
+ optimizer:
81
+ _target_: torch.optim.AdamW
82
+ lr: 1.0e-4
83
+ betas: [0.95, 0.999]
84
+ eps: 1.0e-8
85
+ weight_decay: 1.0e-6
86
+
87
+ training:
88
+ device: "cuda:0"
89
+ seed: 42
90
+ debug: False
91
+ resume: True
92
+ # optimization
93
+ lr_scheduler: cosine
94
+ lr_warmup_steps: 500
95
+ num_epochs: 600
96
+ gradient_accumulate_every: 1
97
+ # EMA destroys performance when used with BatchNorm
98
+ # replace BatchNorm with GroupNorm.
99
+ use_ema: True
100
+ # training loop control
101
+ # in epochs
102
+ rollout_every: 50
103
+ checkpoint_every: 50
104
+ val_every: 1
105
+ sample_every: 5
106
+ # steps per epoch
107
+ max_train_steps: null
108
+ max_val_steps: null
109
+ # misc
110
+ tqdm_interval_sec: 1.0
111
+
112
+ logging:
113
+ project: umi
114
+ resume: True
115
+ mode: online
116
+ name: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
117
+ tags: ["${name}", "${task_name}", "${exp_name}"]
118
+ id: null
119
+ group: null
120
+
121
+ checkpoint:
122
+ topk:
123
+ monitor_key: train_loss
124
+ mode: min
125
+ k: 5
126
+ format_str: 'epoch={epoch:04d}-train_loss={train_loss:.3f}.ckpt'
127
+ save_last_ckpt: True
128
+ save_last_snapshot: False
129
+
130
+ multi_run:
131
+ run_dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
132
+ wandb_name_base: ${now:%Y.%m.%d-%H.%M.%S}_${name}_${task_name}
133
+
134
+ hydra:
135
+ job:
136
+ override_dirname: ${name}
137
+ run:
138
+ dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
139
+ sweep:
140
+ dir: data/outputs/${now:%Y.%m.%d}/${now:%H.%M.%S}_${name}_${task_name}
141
+ subdir: ${hydra.job.num}