Instructions to use lukasskellijs/pi05_ab3_peg8mm_hg_dagger_r2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- LeRobot
How to use lukasskellijs/pi05_ab3_peg8mm_hg_dagger_r2 with LeRobot:
- Notebooks
- Google Colab
- Kaggle
Add files using upload-large-folder tool
Browse files- config.json +110 -0
- model.safetensors +3 -0
- policy_postprocessor.json +38 -0
- policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
- policy_preprocessor.json +109 -0
- policy_preprocessor_step_3_normalizer_processor.safetensors +3 -0
- step_001000/config.json +110 -0
- step_001000/model.safetensors +3 -0
- step_001000/policy_postprocessor.json +38 -0
- step_001000/policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
- step_001000/policy_preprocessor.json +109 -0
- step_001000/policy_preprocessor_step_3_normalizer_processor.safetensors +3 -0
- step_002000/config.json +110 -0
- step_002000/model.safetensors +3 -0
- step_002000/policy_postprocessor.json +38 -0
- step_002000/policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
- step_002000/policy_preprocessor.json +109 -0
- step_002000/policy_preprocessor_step_3_normalizer_processor.safetensors +3 -0
- step_003000/config.json +110 -0
- step_003000/model.safetensors +3 -0
- step_003000/policy_postprocessor.json +38 -0
- step_003000/policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
- step_003000/policy_preprocessor.json +109 -0
- step_003000/policy_preprocessor_step_3_normalizer_processor.safetensors +3 -0
- step_004000/config.json +110 -0
- step_004000/model.safetensors +3 -0
- step_004000/policy_postprocessor.json +38 -0
- step_004000/policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
- step_004000/policy_preprocessor.json +109 -0
- step_004000/policy_preprocessor_step_3_normalizer_processor.safetensors +3 -0
config.json
ADDED
|
@@ -0,0 +1,110 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"type": "pi05",
|
| 3 |
+
"n_obs_steps": 1,
|
| 4 |
+
"input_features": {
|
| 5 |
+
"observation.images.base_0_rgb": {
|
| 6 |
+
"type": "VISUAL",
|
| 7 |
+
"shape": [
|
| 8 |
+
3,
|
| 9 |
+
224,
|
| 10 |
+
224
|
| 11 |
+
]
|
| 12 |
+
},
|
| 13 |
+
"observation.images.left_wrist_0_rgb": {
|
| 14 |
+
"type": "VISUAL",
|
| 15 |
+
"shape": [
|
| 16 |
+
3,
|
| 17 |
+
224,
|
| 18 |
+
224
|
| 19 |
+
]
|
| 20 |
+
},
|
| 21 |
+
"observation.images.right_wrist_0_rgb": {
|
| 22 |
+
"type": "VISUAL",
|
| 23 |
+
"shape": [
|
| 24 |
+
3,
|
| 25 |
+
224,
|
| 26 |
+
224
|
| 27 |
+
]
|
| 28 |
+
},
|
| 29 |
+
"observation.state": {
|
| 30 |
+
"type": "STATE",
|
| 31 |
+
"shape": [
|
| 32 |
+
8
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"output_features": {
|
| 37 |
+
"action": {
|
| 38 |
+
"type": "ACTION",
|
| 39 |
+
"shape": [
|
| 40 |
+
8
|
| 41 |
+
]
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"device": "cuda",
|
| 45 |
+
"use_amp": false,
|
| 46 |
+
"use_peft": false,
|
| 47 |
+
"push_to_hub": false,
|
| 48 |
+
"repo_id": null,
|
| 49 |
+
"private": null,
|
| 50 |
+
"tags": null,
|
| 51 |
+
"license": null,
|
| 52 |
+
"pretrained_path": "/home/ubuntu/outputs/pi05_ab3_peg8mm_hg_dagger_r1/step_004000",
|
| 53 |
+
"pretrained_revision": null,
|
| 54 |
+
"paligemma_variant": "gemma_2b",
|
| 55 |
+
"action_expert_variant": "gemma_300m",
|
| 56 |
+
"dtype": "bfloat16",
|
| 57 |
+
"chunk_size": 15,
|
| 58 |
+
"n_action_steps": 15,
|
| 59 |
+
"max_state_dim": 8,
|
| 60 |
+
"max_action_dim": 32,
|
| 61 |
+
"num_inference_steps": 10,
|
| 62 |
+
"time_sampling_beta_alpha": 1.5,
|
| 63 |
+
"time_sampling_beta_beta": 1.0,
|
| 64 |
+
"time_sampling_scale": 0.999,
|
| 65 |
+
"time_sampling_offset": 0.001,
|
| 66 |
+
"min_period": 0.004,
|
| 67 |
+
"max_period": 4.0,
|
| 68 |
+
"use_relative_actions": true,
|
| 69 |
+
"relative_exclude_joints": [
|
| 70 |
+
"gripper"
|
| 71 |
+
],
|
| 72 |
+
"action_feature_names": [
|
| 73 |
+
"joint_0",
|
| 74 |
+
"joint_1",
|
| 75 |
+
"joint_2",
|
| 76 |
+
"joint_3",
|
| 77 |
+
"joint_4",
|
| 78 |
+
"joint_5",
|
| 79 |
+
"joint_6",
|
| 80 |
+
"gripper"
|
| 81 |
+
],
|
| 82 |
+
"rtc_config": null,
|
| 83 |
+
"image_resolution": [
|
| 84 |
+
224,
|
| 85 |
+
224
|
| 86 |
+
],
|
| 87 |
+
"empty_cameras": 0,
|
| 88 |
+
"tokenizer_max_length": 200,
|
| 89 |
+
"normalization_mapping": {
|
| 90 |
+
"VISUAL": "IDENTITY",
|
| 91 |
+
"STATE": "QUANTILES",
|
| 92 |
+
"ACTION": "QUANTILES"
|
| 93 |
+
},
|
| 94 |
+
"gradient_checkpointing": false,
|
| 95 |
+
"compile_model": false,
|
| 96 |
+
"compile_mode": "max-autotune",
|
| 97 |
+
"freeze_vision_encoder": false,
|
| 98 |
+
"train_expert_only": false,
|
| 99 |
+
"optimizer_lr": 2.5e-05,
|
| 100 |
+
"optimizer_betas": [
|
| 101 |
+
0.9,
|
| 102 |
+
0.95
|
| 103 |
+
],
|
| 104 |
+
"optimizer_eps": 1e-08,
|
| 105 |
+
"optimizer_weight_decay": 0.01,
|
| 106 |
+
"optimizer_grad_clip_norm": 1.0,
|
| 107 |
+
"scheduler_warmup_steps": 1000,
|
| 108 |
+
"scheduler_decay_steps": 30000,
|
| 109 |
+
"scheduler_decay_lr": 2.5e-06
|
| 110 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bcca9e6c36cb4cbcd136d4788c505324a16621562f58864909ce9a522e8963a1
|
| 3 |
+
size 9354050752
|
policy_postprocessor.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_postprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "unnormalizer_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"eps": 1e-08,
|
| 8 |
+
"features": {
|
| 9 |
+
"action": {
|
| 10 |
+
"type": "ACTION",
|
| 11 |
+
"shape": [
|
| 12 |
+
8
|
| 13 |
+
]
|
| 14 |
+
}
|
| 15 |
+
},
|
| 16 |
+
"norm_map": {
|
| 17 |
+
"VISUAL": "IDENTITY",
|
| 18 |
+
"STATE": "QUANTILES",
|
| 19 |
+
"ACTION": "QUANTILES"
|
| 20 |
+
}
|
| 21 |
+
},
|
| 22 |
+
"state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"registry_name": "absolute_actions_processor",
|
| 26 |
+
"config": {
|
| 27 |
+
"enabled": true
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"registry_name": "device_processor",
|
| 32 |
+
"config": {
|
| 33 |
+
"device": "cpu",
|
| 34 |
+
"float_dtype": null
|
| 35 |
+
}
|
| 36 |
+
}
|
| 37 |
+
]
|
| 38 |
+
}
|
policy_postprocessor_step_0_unnormalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50d458bdfa5336a7d87d3addc7ed0777f0ce63e5ec74fe58d4cd5c3ecfe8e0e1
|
| 3 |
+
size 10692
|
policy_preprocessor.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_preprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "rename_observations_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"rename_map": {
|
| 8 |
+
"observation.images.front": "observation.images.base_0_rgb",
|
| 9 |
+
"observation.images.wrist": "observation.images.left_wrist_0_rgb"
|
| 10 |
+
}
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
{
|
| 14 |
+
"registry_name": "to_batch_processor",
|
| 15 |
+
"config": {}
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"registry_name": "relative_actions_processor",
|
| 19 |
+
"config": {
|
| 20 |
+
"enabled": true,
|
| 21 |
+
"exclude_joints": [
|
| 22 |
+
"gripper"
|
| 23 |
+
],
|
| 24 |
+
"action_names": [
|
| 25 |
+
"joint_0",
|
| 26 |
+
"joint_1",
|
| 27 |
+
"joint_2",
|
| 28 |
+
"joint_3",
|
| 29 |
+
"joint_4",
|
| 30 |
+
"joint_5",
|
| 31 |
+
"joint_6",
|
| 32 |
+
"gripper"
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"registry_name": "normalizer_processor",
|
| 38 |
+
"config": {
|
| 39 |
+
"eps": 1e-08,
|
| 40 |
+
"features": {
|
| 41 |
+
"observation.images.base_0_rgb": {
|
| 42 |
+
"type": "VISUAL",
|
| 43 |
+
"shape": [
|
| 44 |
+
3,
|
| 45 |
+
224,
|
| 46 |
+
224
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
"observation.images.left_wrist_0_rgb": {
|
| 50 |
+
"type": "VISUAL",
|
| 51 |
+
"shape": [
|
| 52 |
+
3,
|
| 53 |
+
224,
|
| 54 |
+
224
|
| 55 |
+
]
|
| 56 |
+
},
|
| 57 |
+
"observation.images.right_wrist_0_rgb": {
|
| 58 |
+
"type": "VISUAL",
|
| 59 |
+
"shape": [
|
| 60 |
+
3,
|
| 61 |
+
224,
|
| 62 |
+
224
|
| 63 |
+
]
|
| 64 |
+
},
|
| 65 |
+
"observation.state": {
|
| 66 |
+
"type": "STATE",
|
| 67 |
+
"shape": [
|
| 68 |
+
8
|
| 69 |
+
]
|
| 70 |
+
},
|
| 71 |
+
"action": {
|
| 72 |
+
"type": "ACTION",
|
| 73 |
+
"shape": [
|
| 74 |
+
8
|
| 75 |
+
]
|
| 76 |
+
}
|
| 77 |
+
},
|
| 78 |
+
"norm_map": {
|
| 79 |
+
"VISUAL": "IDENTITY",
|
| 80 |
+
"STATE": "QUANTILES",
|
| 81 |
+
"ACTION": "QUANTILES"
|
| 82 |
+
}
|
| 83 |
+
},
|
| 84 |
+
"state_file": "policy_preprocessor_step_3_normalizer_processor.safetensors"
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"registry_name": "pi05_prepare_state_tokenizer_processor_step",
|
| 88 |
+
"config": {}
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"registry_name": "tokenizer_processor",
|
| 92 |
+
"config": {
|
| 93 |
+
"max_length": 200,
|
| 94 |
+
"task_key": "task",
|
| 95 |
+
"padding_side": "right",
|
| 96 |
+
"padding": "max_length",
|
| 97 |
+
"truncation": true,
|
| 98 |
+
"tokenizer_name": "google/paligemma-3b-pt-224"
|
| 99 |
+
}
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"registry_name": "device_processor",
|
| 103 |
+
"config": {
|
| 104 |
+
"device": "cuda",
|
| 105 |
+
"float_dtype": null
|
| 106 |
+
}
|
| 107 |
+
}
|
| 108 |
+
]
|
| 109 |
+
}
|
policy_preprocessor_step_3_normalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50d458bdfa5336a7d87d3addc7ed0777f0ce63e5ec74fe58d4cd5c3ecfe8e0e1
|
| 3 |
+
size 10692
|
step_001000/config.json
ADDED
|
@@ -0,0 +1,110 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"type": "pi05",
|
| 3 |
+
"n_obs_steps": 1,
|
| 4 |
+
"input_features": {
|
| 5 |
+
"observation.images.base_0_rgb": {
|
| 6 |
+
"type": "VISUAL",
|
| 7 |
+
"shape": [
|
| 8 |
+
3,
|
| 9 |
+
224,
|
| 10 |
+
224
|
| 11 |
+
]
|
| 12 |
+
},
|
| 13 |
+
"observation.images.left_wrist_0_rgb": {
|
| 14 |
+
"type": "VISUAL",
|
| 15 |
+
"shape": [
|
| 16 |
+
3,
|
| 17 |
+
224,
|
| 18 |
+
224
|
| 19 |
+
]
|
| 20 |
+
},
|
| 21 |
+
"observation.images.right_wrist_0_rgb": {
|
| 22 |
+
"type": "VISUAL",
|
| 23 |
+
"shape": [
|
| 24 |
+
3,
|
| 25 |
+
224,
|
| 26 |
+
224
|
| 27 |
+
]
|
| 28 |
+
},
|
| 29 |
+
"observation.state": {
|
| 30 |
+
"type": "STATE",
|
| 31 |
+
"shape": [
|
| 32 |
+
8
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"output_features": {
|
| 37 |
+
"action": {
|
| 38 |
+
"type": "ACTION",
|
| 39 |
+
"shape": [
|
| 40 |
+
8
|
| 41 |
+
]
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"device": "cuda",
|
| 45 |
+
"use_amp": false,
|
| 46 |
+
"use_peft": false,
|
| 47 |
+
"push_to_hub": false,
|
| 48 |
+
"repo_id": null,
|
| 49 |
+
"private": null,
|
| 50 |
+
"tags": null,
|
| 51 |
+
"license": null,
|
| 52 |
+
"pretrained_path": "/home/ubuntu/outputs/pi05_ab3_peg8mm_hg_dagger_r1/step_004000",
|
| 53 |
+
"pretrained_revision": null,
|
| 54 |
+
"paligemma_variant": "gemma_2b",
|
| 55 |
+
"action_expert_variant": "gemma_300m",
|
| 56 |
+
"dtype": "bfloat16",
|
| 57 |
+
"chunk_size": 15,
|
| 58 |
+
"n_action_steps": 15,
|
| 59 |
+
"max_state_dim": 8,
|
| 60 |
+
"max_action_dim": 32,
|
| 61 |
+
"num_inference_steps": 10,
|
| 62 |
+
"time_sampling_beta_alpha": 1.5,
|
| 63 |
+
"time_sampling_beta_beta": 1.0,
|
| 64 |
+
"time_sampling_scale": 0.999,
|
| 65 |
+
"time_sampling_offset": 0.001,
|
| 66 |
+
"min_period": 0.004,
|
| 67 |
+
"max_period": 4.0,
|
| 68 |
+
"use_relative_actions": true,
|
| 69 |
+
"relative_exclude_joints": [
|
| 70 |
+
"gripper"
|
| 71 |
+
],
|
| 72 |
+
"action_feature_names": [
|
| 73 |
+
"joint_0",
|
| 74 |
+
"joint_1",
|
| 75 |
+
"joint_2",
|
| 76 |
+
"joint_3",
|
| 77 |
+
"joint_4",
|
| 78 |
+
"joint_5",
|
| 79 |
+
"joint_6",
|
| 80 |
+
"gripper"
|
| 81 |
+
],
|
| 82 |
+
"rtc_config": null,
|
| 83 |
+
"image_resolution": [
|
| 84 |
+
224,
|
| 85 |
+
224
|
| 86 |
+
],
|
| 87 |
+
"empty_cameras": 0,
|
| 88 |
+
"tokenizer_max_length": 200,
|
| 89 |
+
"normalization_mapping": {
|
| 90 |
+
"VISUAL": "IDENTITY",
|
| 91 |
+
"STATE": "QUANTILES",
|
| 92 |
+
"ACTION": "QUANTILES"
|
| 93 |
+
},
|
| 94 |
+
"gradient_checkpointing": false,
|
| 95 |
+
"compile_model": false,
|
| 96 |
+
"compile_mode": "max-autotune",
|
| 97 |
+
"freeze_vision_encoder": false,
|
| 98 |
+
"train_expert_only": false,
|
| 99 |
+
"optimizer_lr": 2.5e-05,
|
| 100 |
+
"optimizer_betas": [
|
| 101 |
+
0.9,
|
| 102 |
+
0.95
|
| 103 |
+
],
|
| 104 |
+
"optimizer_eps": 1e-08,
|
| 105 |
+
"optimizer_weight_decay": 0.01,
|
| 106 |
+
"optimizer_grad_clip_norm": 1.0,
|
| 107 |
+
"scheduler_warmup_steps": 1000,
|
| 108 |
+
"scheduler_decay_steps": 30000,
|
| 109 |
+
"scheduler_decay_lr": 2.5e-06
|
| 110 |
+
}
|
step_001000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:acf448a4d276ff31874e2e9ab3235178e16218b2ce9050ac44a7a3418dd9a976
|
| 3 |
+
size 9354050752
|
step_001000/policy_postprocessor.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_postprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "unnormalizer_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"eps": 1e-08,
|
| 8 |
+
"features": {
|
| 9 |
+
"action": {
|
| 10 |
+
"type": "ACTION",
|
| 11 |
+
"shape": [
|
| 12 |
+
8
|
| 13 |
+
]
|
| 14 |
+
}
|
| 15 |
+
},
|
| 16 |
+
"norm_map": {
|
| 17 |
+
"VISUAL": "IDENTITY",
|
| 18 |
+
"STATE": "QUANTILES",
|
| 19 |
+
"ACTION": "QUANTILES"
|
| 20 |
+
}
|
| 21 |
+
},
|
| 22 |
+
"state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"registry_name": "absolute_actions_processor",
|
| 26 |
+
"config": {
|
| 27 |
+
"enabled": true
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"registry_name": "device_processor",
|
| 32 |
+
"config": {
|
| 33 |
+
"device": "cpu",
|
| 34 |
+
"float_dtype": null
|
| 35 |
+
}
|
| 36 |
+
}
|
| 37 |
+
]
|
| 38 |
+
}
|
step_001000/policy_postprocessor_step_0_unnormalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50d458bdfa5336a7d87d3addc7ed0777f0ce63e5ec74fe58d4cd5c3ecfe8e0e1
|
| 3 |
+
size 10692
|
step_001000/policy_preprocessor.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_preprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "rename_observations_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"rename_map": {
|
| 8 |
+
"observation.images.front": "observation.images.base_0_rgb",
|
| 9 |
+
"observation.images.wrist": "observation.images.left_wrist_0_rgb"
|
| 10 |
+
}
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
{
|
| 14 |
+
"registry_name": "to_batch_processor",
|
| 15 |
+
"config": {}
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"registry_name": "relative_actions_processor",
|
| 19 |
+
"config": {
|
| 20 |
+
"enabled": true,
|
| 21 |
+
"exclude_joints": [
|
| 22 |
+
"gripper"
|
| 23 |
+
],
|
| 24 |
+
"action_names": [
|
| 25 |
+
"joint_0",
|
| 26 |
+
"joint_1",
|
| 27 |
+
"joint_2",
|
| 28 |
+
"joint_3",
|
| 29 |
+
"joint_4",
|
| 30 |
+
"joint_5",
|
| 31 |
+
"joint_6",
|
| 32 |
+
"gripper"
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"registry_name": "normalizer_processor",
|
| 38 |
+
"config": {
|
| 39 |
+
"eps": 1e-08,
|
| 40 |
+
"features": {
|
| 41 |
+
"observation.images.base_0_rgb": {
|
| 42 |
+
"type": "VISUAL",
|
| 43 |
+
"shape": [
|
| 44 |
+
3,
|
| 45 |
+
224,
|
| 46 |
+
224
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
"observation.images.left_wrist_0_rgb": {
|
| 50 |
+
"type": "VISUAL",
|
| 51 |
+
"shape": [
|
| 52 |
+
3,
|
| 53 |
+
224,
|
| 54 |
+
224
|
| 55 |
+
]
|
| 56 |
+
},
|
| 57 |
+
"observation.images.right_wrist_0_rgb": {
|
| 58 |
+
"type": "VISUAL",
|
| 59 |
+
"shape": [
|
| 60 |
+
3,
|
| 61 |
+
224,
|
| 62 |
+
224
|
| 63 |
+
]
|
| 64 |
+
},
|
| 65 |
+
"observation.state": {
|
| 66 |
+
"type": "STATE",
|
| 67 |
+
"shape": [
|
| 68 |
+
8
|
| 69 |
+
]
|
| 70 |
+
},
|
| 71 |
+
"action": {
|
| 72 |
+
"type": "ACTION",
|
| 73 |
+
"shape": [
|
| 74 |
+
8
|
| 75 |
+
]
|
| 76 |
+
}
|
| 77 |
+
},
|
| 78 |
+
"norm_map": {
|
| 79 |
+
"VISUAL": "IDENTITY",
|
| 80 |
+
"STATE": "QUANTILES",
|
| 81 |
+
"ACTION": "QUANTILES"
|
| 82 |
+
}
|
| 83 |
+
},
|
| 84 |
+
"state_file": "policy_preprocessor_step_3_normalizer_processor.safetensors"
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"registry_name": "pi05_prepare_state_tokenizer_processor_step",
|
| 88 |
+
"config": {}
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"registry_name": "tokenizer_processor",
|
| 92 |
+
"config": {
|
| 93 |
+
"max_length": 200,
|
| 94 |
+
"task_key": "task",
|
| 95 |
+
"padding_side": "right",
|
| 96 |
+
"padding": "max_length",
|
| 97 |
+
"truncation": true,
|
| 98 |
+
"tokenizer_name": "google/paligemma-3b-pt-224"
|
| 99 |
+
}
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"registry_name": "device_processor",
|
| 103 |
+
"config": {
|
| 104 |
+
"device": "cuda",
|
| 105 |
+
"float_dtype": null
|
| 106 |
+
}
|
| 107 |
+
}
|
| 108 |
+
]
|
| 109 |
+
}
|
step_001000/policy_preprocessor_step_3_normalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50d458bdfa5336a7d87d3addc7ed0777f0ce63e5ec74fe58d4cd5c3ecfe8e0e1
|
| 3 |
+
size 10692
|
step_002000/config.json
ADDED
|
@@ -0,0 +1,110 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"type": "pi05",
|
| 3 |
+
"n_obs_steps": 1,
|
| 4 |
+
"input_features": {
|
| 5 |
+
"observation.images.base_0_rgb": {
|
| 6 |
+
"type": "VISUAL",
|
| 7 |
+
"shape": [
|
| 8 |
+
3,
|
| 9 |
+
224,
|
| 10 |
+
224
|
| 11 |
+
]
|
| 12 |
+
},
|
| 13 |
+
"observation.images.left_wrist_0_rgb": {
|
| 14 |
+
"type": "VISUAL",
|
| 15 |
+
"shape": [
|
| 16 |
+
3,
|
| 17 |
+
224,
|
| 18 |
+
224
|
| 19 |
+
]
|
| 20 |
+
},
|
| 21 |
+
"observation.images.right_wrist_0_rgb": {
|
| 22 |
+
"type": "VISUAL",
|
| 23 |
+
"shape": [
|
| 24 |
+
3,
|
| 25 |
+
224,
|
| 26 |
+
224
|
| 27 |
+
]
|
| 28 |
+
},
|
| 29 |
+
"observation.state": {
|
| 30 |
+
"type": "STATE",
|
| 31 |
+
"shape": [
|
| 32 |
+
8
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"output_features": {
|
| 37 |
+
"action": {
|
| 38 |
+
"type": "ACTION",
|
| 39 |
+
"shape": [
|
| 40 |
+
8
|
| 41 |
+
]
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"device": "cuda",
|
| 45 |
+
"use_amp": false,
|
| 46 |
+
"use_peft": false,
|
| 47 |
+
"push_to_hub": false,
|
| 48 |
+
"repo_id": null,
|
| 49 |
+
"private": null,
|
| 50 |
+
"tags": null,
|
| 51 |
+
"license": null,
|
| 52 |
+
"pretrained_path": "/home/ubuntu/outputs/pi05_ab3_peg8mm_hg_dagger_r1/step_004000",
|
| 53 |
+
"pretrained_revision": null,
|
| 54 |
+
"paligemma_variant": "gemma_2b",
|
| 55 |
+
"action_expert_variant": "gemma_300m",
|
| 56 |
+
"dtype": "bfloat16",
|
| 57 |
+
"chunk_size": 15,
|
| 58 |
+
"n_action_steps": 15,
|
| 59 |
+
"max_state_dim": 8,
|
| 60 |
+
"max_action_dim": 32,
|
| 61 |
+
"num_inference_steps": 10,
|
| 62 |
+
"time_sampling_beta_alpha": 1.5,
|
| 63 |
+
"time_sampling_beta_beta": 1.0,
|
| 64 |
+
"time_sampling_scale": 0.999,
|
| 65 |
+
"time_sampling_offset": 0.001,
|
| 66 |
+
"min_period": 0.004,
|
| 67 |
+
"max_period": 4.0,
|
| 68 |
+
"use_relative_actions": true,
|
| 69 |
+
"relative_exclude_joints": [
|
| 70 |
+
"gripper"
|
| 71 |
+
],
|
| 72 |
+
"action_feature_names": [
|
| 73 |
+
"joint_0",
|
| 74 |
+
"joint_1",
|
| 75 |
+
"joint_2",
|
| 76 |
+
"joint_3",
|
| 77 |
+
"joint_4",
|
| 78 |
+
"joint_5",
|
| 79 |
+
"joint_6",
|
| 80 |
+
"gripper"
|
| 81 |
+
],
|
| 82 |
+
"rtc_config": null,
|
| 83 |
+
"image_resolution": [
|
| 84 |
+
224,
|
| 85 |
+
224
|
| 86 |
+
],
|
| 87 |
+
"empty_cameras": 0,
|
| 88 |
+
"tokenizer_max_length": 200,
|
| 89 |
+
"normalization_mapping": {
|
| 90 |
+
"VISUAL": "IDENTITY",
|
| 91 |
+
"STATE": "QUANTILES",
|
| 92 |
+
"ACTION": "QUANTILES"
|
| 93 |
+
},
|
| 94 |
+
"gradient_checkpointing": false,
|
| 95 |
+
"compile_model": false,
|
| 96 |
+
"compile_mode": "max-autotune",
|
| 97 |
+
"freeze_vision_encoder": false,
|
| 98 |
+
"train_expert_only": false,
|
| 99 |
+
"optimizer_lr": 2.5e-05,
|
| 100 |
+
"optimizer_betas": [
|
| 101 |
+
0.9,
|
| 102 |
+
0.95
|
| 103 |
+
],
|
| 104 |
+
"optimizer_eps": 1e-08,
|
| 105 |
+
"optimizer_weight_decay": 0.01,
|
| 106 |
+
"optimizer_grad_clip_norm": 1.0,
|
| 107 |
+
"scheduler_warmup_steps": 1000,
|
| 108 |
+
"scheduler_decay_steps": 30000,
|
| 109 |
+
"scheduler_decay_lr": 2.5e-06
|
| 110 |
+
}
|
step_002000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:101a19875a2f1e3f1f2966fe296fd456d6eee078393e3572793f1f61d8c63cfa
|
| 3 |
+
size 9354050752
|
step_002000/policy_postprocessor.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_postprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "unnormalizer_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"eps": 1e-08,
|
| 8 |
+
"features": {
|
| 9 |
+
"action": {
|
| 10 |
+
"type": "ACTION",
|
| 11 |
+
"shape": [
|
| 12 |
+
8
|
| 13 |
+
]
|
| 14 |
+
}
|
| 15 |
+
},
|
| 16 |
+
"norm_map": {
|
| 17 |
+
"VISUAL": "IDENTITY",
|
| 18 |
+
"STATE": "QUANTILES",
|
| 19 |
+
"ACTION": "QUANTILES"
|
| 20 |
+
}
|
| 21 |
+
},
|
| 22 |
+
"state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"registry_name": "absolute_actions_processor",
|
| 26 |
+
"config": {
|
| 27 |
+
"enabled": true
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"registry_name": "device_processor",
|
| 32 |
+
"config": {
|
| 33 |
+
"device": "cpu",
|
| 34 |
+
"float_dtype": null
|
| 35 |
+
}
|
| 36 |
+
}
|
| 37 |
+
]
|
| 38 |
+
}
|
step_002000/policy_postprocessor_step_0_unnormalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50d458bdfa5336a7d87d3addc7ed0777f0ce63e5ec74fe58d4cd5c3ecfe8e0e1
|
| 3 |
+
size 10692
|
step_002000/policy_preprocessor.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_preprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "rename_observations_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"rename_map": {
|
| 8 |
+
"observation.images.front": "observation.images.base_0_rgb",
|
| 9 |
+
"observation.images.wrist": "observation.images.left_wrist_0_rgb"
|
| 10 |
+
}
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
{
|
| 14 |
+
"registry_name": "to_batch_processor",
|
| 15 |
+
"config": {}
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"registry_name": "relative_actions_processor",
|
| 19 |
+
"config": {
|
| 20 |
+
"enabled": true,
|
| 21 |
+
"exclude_joints": [
|
| 22 |
+
"gripper"
|
| 23 |
+
],
|
| 24 |
+
"action_names": [
|
| 25 |
+
"joint_0",
|
| 26 |
+
"joint_1",
|
| 27 |
+
"joint_2",
|
| 28 |
+
"joint_3",
|
| 29 |
+
"joint_4",
|
| 30 |
+
"joint_5",
|
| 31 |
+
"joint_6",
|
| 32 |
+
"gripper"
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"registry_name": "normalizer_processor",
|
| 38 |
+
"config": {
|
| 39 |
+
"eps": 1e-08,
|
| 40 |
+
"features": {
|
| 41 |
+
"observation.images.base_0_rgb": {
|
| 42 |
+
"type": "VISUAL",
|
| 43 |
+
"shape": [
|
| 44 |
+
3,
|
| 45 |
+
224,
|
| 46 |
+
224
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
"observation.images.left_wrist_0_rgb": {
|
| 50 |
+
"type": "VISUAL",
|
| 51 |
+
"shape": [
|
| 52 |
+
3,
|
| 53 |
+
224,
|
| 54 |
+
224
|
| 55 |
+
]
|
| 56 |
+
},
|
| 57 |
+
"observation.images.right_wrist_0_rgb": {
|
| 58 |
+
"type": "VISUAL",
|
| 59 |
+
"shape": [
|
| 60 |
+
3,
|
| 61 |
+
224,
|
| 62 |
+
224
|
| 63 |
+
]
|
| 64 |
+
},
|
| 65 |
+
"observation.state": {
|
| 66 |
+
"type": "STATE",
|
| 67 |
+
"shape": [
|
| 68 |
+
8
|
| 69 |
+
]
|
| 70 |
+
},
|
| 71 |
+
"action": {
|
| 72 |
+
"type": "ACTION",
|
| 73 |
+
"shape": [
|
| 74 |
+
8
|
| 75 |
+
]
|
| 76 |
+
}
|
| 77 |
+
},
|
| 78 |
+
"norm_map": {
|
| 79 |
+
"VISUAL": "IDENTITY",
|
| 80 |
+
"STATE": "QUANTILES",
|
| 81 |
+
"ACTION": "QUANTILES"
|
| 82 |
+
}
|
| 83 |
+
},
|
| 84 |
+
"state_file": "policy_preprocessor_step_3_normalizer_processor.safetensors"
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"registry_name": "pi05_prepare_state_tokenizer_processor_step",
|
| 88 |
+
"config": {}
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"registry_name": "tokenizer_processor",
|
| 92 |
+
"config": {
|
| 93 |
+
"max_length": 200,
|
| 94 |
+
"task_key": "task",
|
| 95 |
+
"padding_side": "right",
|
| 96 |
+
"padding": "max_length",
|
| 97 |
+
"truncation": true,
|
| 98 |
+
"tokenizer_name": "google/paligemma-3b-pt-224"
|
| 99 |
+
}
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"registry_name": "device_processor",
|
| 103 |
+
"config": {
|
| 104 |
+
"device": "cuda",
|
| 105 |
+
"float_dtype": null
|
| 106 |
+
}
|
| 107 |
+
}
|
| 108 |
+
]
|
| 109 |
+
}
|
step_002000/policy_preprocessor_step_3_normalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50d458bdfa5336a7d87d3addc7ed0777f0ce63e5ec74fe58d4cd5c3ecfe8e0e1
|
| 3 |
+
size 10692
|
step_003000/config.json
ADDED
|
@@ -0,0 +1,110 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"type": "pi05",
|
| 3 |
+
"n_obs_steps": 1,
|
| 4 |
+
"input_features": {
|
| 5 |
+
"observation.images.base_0_rgb": {
|
| 6 |
+
"type": "VISUAL",
|
| 7 |
+
"shape": [
|
| 8 |
+
3,
|
| 9 |
+
224,
|
| 10 |
+
224
|
| 11 |
+
]
|
| 12 |
+
},
|
| 13 |
+
"observation.images.left_wrist_0_rgb": {
|
| 14 |
+
"type": "VISUAL",
|
| 15 |
+
"shape": [
|
| 16 |
+
3,
|
| 17 |
+
224,
|
| 18 |
+
224
|
| 19 |
+
]
|
| 20 |
+
},
|
| 21 |
+
"observation.images.right_wrist_0_rgb": {
|
| 22 |
+
"type": "VISUAL",
|
| 23 |
+
"shape": [
|
| 24 |
+
3,
|
| 25 |
+
224,
|
| 26 |
+
224
|
| 27 |
+
]
|
| 28 |
+
},
|
| 29 |
+
"observation.state": {
|
| 30 |
+
"type": "STATE",
|
| 31 |
+
"shape": [
|
| 32 |
+
8
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"output_features": {
|
| 37 |
+
"action": {
|
| 38 |
+
"type": "ACTION",
|
| 39 |
+
"shape": [
|
| 40 |
+
8
|
| 41 |
+
]
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"device": "cuda",
|
| 45 |
+
"use_amp": false,
|
| 46 |
+
"use_peft": false,
|
| 47 |
+
"push_to_hub": false,
|
| 48 |
+
"repo_id": null,
|
| 49 |
+
"private": null,
|
| 50 |
+
"tags": null,
|
| 51 |
+
"license": null,
|
| 52 |
+
"pretrained_path": "/home/ubuntu/outputs/pi05_ab3_peg8mm_hg_dagger_r1/step_004000",
|
| 53 |
+
"pretrained_revision": null,
|
| 54 |
+
"paligemma_variant": "gemma_2b",
|
| 55 |
+
"action_expert_variant": "gemma_300m",
|
| 56 |
+
"dtype": "bfloat16",
|
| 57 |
+
"chunk_size": 15,
|
| 58 |
+
"n_action_steps": 15,
|
| 59 |
+
"max_state_dim": 8,
|
| 60 |
+
"max_action_dim": 32,
|
| 61 |
+
"num_inference_steps": 10,
|
| 62 |
+
"time_sampling_beta_alpha": 1.5,
|
| 63 |
+
"time_sampling_beta_beta": 1.0,
|
| 64 |
+
"time_sampling_scale": 0.999,
|
| 65 |
+
"time_sampling_offset": 0.001,
|
| 66 |
+
"min_period": 0.004,
|
| 67 |
+
"max_period": 4.0,
|
| 68 |
+
"use_relative_actions": true,
|
| 69 |
+
"relative_exclude_joints": [
|
| 70 |
+
"gripper"
|
| 71 |
+
],
|
| 72 |
+
"action_feature_names": [
|
| 73 |
+
"joint_0",
|
| 74 |
+
"joint_1",
|
| 75 |
+
"joint_2",
|
| 76 |
+
"joint_3",
|
| 77 |
+
"joint_4",
|
| 78 |
+
"joint_5",
|
| 79 |
+
"joint_6",
|
| 80 |
+
"gripper"
|
| 81 |
+
],
|
| 82 |
+
"rtc_config": null,
|
| 83 |
+
"image_resolution": [
|
| 84 |
+
224,
|
| 85 |
+
224
|
| 86 |
+
],
|
| 87 |
+
"empty_cameras": 0,
|
| 88 |
+
"tokenizer_max_length": 200,
|
| 89 |
+
"normalization_mapping": {
|
| 90 |
+
"VISUAL": "IDENTITY",
|
| 91 |
+
"STATE": "QUANTILES",
|
| 92 |
+
"ACTION": "QUANTILES"
|
| 93 |
+
},
|
| 94 |
+
"gradient_checkpointing": false,
|
| 95 |
+
"compile_model": false,
|
| 96 |
+
"compile_mode": "max-autotune",
|
| 97 |
+
"freeze_vision_encoder": false,
|
| 98 |
+
"train_expert_only": false,
|
| 99 |
+
"optimizer_lr": 2.5e-05,
|
| 100 |
+
"optimizer_betas": [
|
| 101 |
+
0.9,
|
| 102 |
+
0.95
|
| 103 |
+
],
|
| 104 |
+
"optimizer_eps": 1e-08,
|
| 105 |
+
"optimizer_weight_decay": 0.01,
|
| 106 |
+
"optimizer_grad_clip_norm": 1.0,
|
| 107 |
+
"scheduler_warmup_steps": 1000,
|
| 108 |
+
"scheduler_decay_steps": 30000,
|
| 109 |
+
"scheduler_decay_lr": 2.5e-06
|
| 110 |
+
}
|
step_003000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d11c514f7dff5eb296ade6e800a27ef978ffa63109dbf2a17cc48ccc77fe54b
|
| 3 |
+
size 9354050752
|
step_003000/policy_postprocessor.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_postprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "unnormalizer_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"eps": 1e-08,
|
| 8 |
+
"features": {
|
| 9 |
+
"action": {
|
| 10 |
+
"type": "ACTION",
|
| 11 |
+
"shape": [
|
| 12 |
+
8
|
| 13 |
+
]
|
| 14 |
+
}
|
| 15 |
+
},
|
| 16 |
+
"norm_map": {
|
| 17 |
+
"VISUAL": "IDENTITY",
|
| 18 |
+
"STATE": "QUANTILES",
|
| 19 |
+
"ACTION": "QUANTILES"
|
| 20 |
+
}
|
| 21 |
+
},
|
| 22 |
+
"state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"registry_name": "absolute_actions_processor",
|
| 26 |
+
"config": {
|
| 27 |
+
"enabled": true
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"registry_name": "device_processor",
|
| 32 |
+
"config": {
|
| 33 |
+
"device": "cpu",
|
| 34 |
+
"float_dtype": null
|
| 35 |
+
}
|
| 36 |
+
}
|
| 37 |
+
]
|
| 38 |
+
}
|
step_003000/policy_postprocessor_step_0_unnormalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50d458bdfa5336a7d87d3addc7ed0777f0ce63e5ec74fe58d4cd5c3ecfe8e0e1
|
| 3 |
+
size 10692
|
step_003000/policy_preprocessor.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_preprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "rename_observations_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"rename_map": {
|
| 8 |
+
"observation.images.front": "observation.images.base_0_rgb",
|
| 9 |
+
"observation.images.wrist": "observation.images.left_wrist_0_rgb"
|
| 10 |
+
}
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
{
|
| 14 |
+
"registry_name": "to_batch_processor",
|
| 15 |
+
"config": {}
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"registry_name": "relative_actions_processor",
|
| 19 |
+
"config": {
|
| 20 |
+
"enabled": true,
|
| 21 |
+
"exclude_joints": [
|
| 22 |
+
"gripper"
|
| 23 |
+
],
|
| 24 |
+
"action_names": [
|
| 25 |
+
"joint_0",
|
| 26 |
+
"joint_1",
|
| 27 |
+
"joint_2",
|
| 28 |
+
"joint_3",
|
| 29 |
+
"joint_4",
|
| 30 |
+
"joint_5",
|
| 31 |
+
"joint_6",
|
| 32 |
+
"gripper"
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"registry_name": "normalizer_processor",
|
| 38 |
+
"config": {
|
| 39 |
+
"eps": 1e-08,
|
| 40 |
+
"features": {
|
| 41 |
+
"observation.images.base_0_rgb": {
|
| 42 |
+
"type": "VISUAL",
|
| 43 |
+
"shape": [
|
| 44 |
+
3,
|
| 45 |
+
224,
|
| 46 |
+
224
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
"observation.images.left_wrist_0_rgb": {
|
| 50 |
+
"type": "VISUAL",
|
| 51 |
+
"shape": [
|
| 52 |
+
3,
|
| 53 |
+
224,
|
| 54 |
+
224
|
| 55 |
+
]
|
| 56 |
+
},
|
| 57 |
+
"observation.images.right_wrist_0_rgb": {
|
| 58 |
+
"type": "VISUAL",
|
| 59 |
+
"shape": [
|
| 60 |
+
3,
|
| 61 |
+
224,
|
| 62 |
+
224
|
| 63 |
+
]
|
| 64 |
+
},
|
| 65 |
+
"observation.state": {
|
| 66 |
+
"type": "STATE",
|
| 67 |
+
"shape": [
|
| 68 |
+
8
|
| 69 |
+
]
|
| 70 |
+
},
|
| 71 |
+
"action": {
|
| 72 |
+
"type": "ACTION",
|
| 73 |
+
"shape": [
|
| 74 |
+
8
|
| 75 |
+
]
|
| 76 |
+
}
|
| 77 |
+
},
|
| 78 |
+
"norm_map": {
|
| 79 |
+
"VISUAL": "IDENTITY",
|
| 80 |
+
"STATE": "QUANTILES",
|
| 81 |
+
"ACTION": "QUANTILES"
|
| 82 |
+
}
|
| 83 |
+
},
|
| 84 |
+
"state_file": "policy_preprocessor_step_3_normalizer_processor.safetensors"
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"registry_name": "pi05_prepare_state_tokenizer_processor_step",
|
| 88 |
+
"config": {}
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"registry_name": "tokenizer_processor",
|
| 92 |
+
"config": {
|
| 93 |
+
"max_length": 200,
|
| 94 |
+
"task_key": "task",
|
| 95 |
+
"padding_side": "right",
|
| 96 |
+
"padding": "max_length",
|
| 97 |
+
"truncation": true,
|
| 98 |
+
"tokenizer_name": "google/paligemma-3b-pt-224"
|
| 99 |
+
}
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"registry_name": "device_processor",
|
| 103 |
+
"config": {
|
| 104 |
+
"device": "cuda",
|
| 105 |
+
"float_dtype": null
|
| 106 |
+
}
|
| 107 |
+
}
|
| 108 |
+
]
|
| 109 |
+
}
|
step_003000/policy_preprocessor_step_3_normalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50d458bdfa5336a7d87d3addc7ed0777f0ce63e5ec74fe58d4cd5c3ecfe8e0e1
|
| 3 |
+
size 10692
|
step_004000/config.json
ADDED
|
@@ -0,0 +1,110 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"type": "pi05",
|
| 3 |
+
"n_obs_steps": 1,
|
| 4 |
+
"input_features": {
|
| 5 |
+
"observation.images.base_0_rgb": {
|
| 6 |
+
"type": "VISUAL",
|
| 7 |
+
"shape": [
|
| 8 |
+
3,
|
| 9 |
+
224,
|
| 10 |
+
224
|
| 11 |
+
]
|
| 12 |
+
},
|
| 13 |
+
"observation.images.left_wrist_0_rgb": {
|
| 14 |
+
"type": "VISUAL",
|
| 15 |
+
"shape": [
|
| 16 |
+
3,
|
| 17 |
+
224,
|
| 18 |
+
224
|
| 19 |
+
]
|
| 20 |
+
},
|
| 21 |
+
"observation.images.right_wrist_0_rgb": {
|
| 22 |
+
"type": "VISUAL",
|
| 23 |
+
"shape": [
|
| 24 |
+
3,
|
| 25 |
+
224,
|
| 26 |
+
224
|
| 27 |
+
]
|
| 28 |
+
},
|
| 29 |
+
"observation.state": {
|
| 30 |
+
"type": "STATE",
|
| 31 |
+
"shape": [
|
| 32 |
+
8
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"output_features": {
|
| 37 |
+
"action": {
|
| 38 |
+
"type": "ACTION",
|
| 39 |
+
"shape": [
|
| 40 |
+
8
|
| 41 |
+
]
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"device": "cuda",
|
| 45 |
+
"use_amp": false,
|
| 46 |
+
"use_peft": false,
|
| 47 |
+
"push_to_hub": false,
|
| 48 |
+
"repo_id": null,
|
| 49 |
+
"private": null,
|
| 50 |
+
"tags": null,
|
| 51 |
+
"license": null,
|
| 52 |
+
"pretrained_path": "/home/ubuntu/outputs/pi05_ab3_peg8mm_hg_dagger_r1/step_004000",
|
| 53 |
+
"pretrained_revision": null,
|
| 54 |
+
"paligemma_variant": "gemma_2b",
|
| 55 |
+
"action_expert_variant": "gemma_300m",
|
| 56 |
+
"dtype": "bfloat16",
|
| 57 |
+
"chunk_size": 15,
|
| 58 |
+
"n_action_steps": 15,
|
| 59 |
+
"max_state_dim": 8,
|
| 60 |
+
"max_action_dim": 32,
|
| 61 |
+
"num_inference_steps": 10,
|
| 62 |
+
"time_sampling_beta_alpha": 1.5,
|
| 63 |
+
"time_sampling_beta_beta": 1.0,
|
| 64 |
+
"time_sampling_scale": 0.999,
|
| 65 |
+
"time_sampling_offset": 0.001,
|
| 66 |
+
"min_period": 0.004,
|
| 67 |
+
"max_period": 4.0,
|
| 68 |
+
"use_relative_actions": true,
|
| 69 |
+
"relative_exclude_joints": [
|
| 70 |
+
"gripper"
|
| 71 |
+
],
|
| 72 |
+
"action_feature_names": [
|
| 73 |
+
"joint_0",
|
| 74 |
+
"joint_1",
|
| 75 |
+
"joint_2",
|
| 76 |
+
"joint_3",
|
| 77 |
+
"joint_4",
|
| 78 |
+
"joint_5",
|
| 79 |
+
"joint_6",
|
| 80 |
+
"gripper"
|
| 81 |
+
],
|
| 82 |
+
"rtc_config": null,
|
| 83 |
+
"image_resolution": [
|
| 84 |
+
224,
|
| 85 |
+
224
|
| 86 |
+
],
|
| 87 |
+
"empty_cameras": 0,
|
| 88 |
+
"tokenizer_max_length": 200,
|
| 89 |
+
"normalization_mapping": {
|
| 90 |
+
"VISUAL": "IDENTITY",
|
| 91 |
+
"STATE": "QUANTILES",
|
| 92 |
+
"ACTION": "QUANTILES"
|
| 93 |
+
},
|
| 94 |
+
"gradient_checkpointing": false,
|
| 95 |
+
"compile_model": false,
|
| 96 |
+
"compile_mode": "max-autotune",
|
| 97 |
+
"freeze_vision_encoder": false,
|
| 98 |
+
"train_expert_only": false,
|
| 99 |
+
"optimizer_lr": 2.5e-05,
|
| 100 |
+
"optimizer_betas": [
|
| 101 |
+
0.9,
|
| 102 |
+
0.95
|
| 103 |
+
],
|
| 104 |
+
"optimizer_eps": 1e-08,
|
| 105 |
+
"optimizer_weight_decay": 0.01,
|
| 106 |
+
"optimizer_grad_clip_norm": 1.0,
|
| 107 |
+
"scheduler_warmup_steps": 1000,
|
| 108 |
+
"scheduler_decay_steps": 30000,
|
| 109 |
+
"scheduler_decay_lr": 2.5e-06
|
| 110 |
+
}
|
step_004000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bcca9e6c36cb4cbcd136d4788c505324a16621562f58864909ce9a522e8963a1
|
| 3 |
+
size 9354050752
|
step_004000/policy_postprocessor.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_postprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "unnormalizer_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"eps": 1e-08,
|
| 8 |
+
"features": {
|
| 9 |
+
"action": {
|
| 10 |
+
"type": "ACTION",
|
| 11 |
+
"shape": [
|
| 12 |
+
8
|
| 13 |
+
]
|
| 14 |
+
}
|
| 15 |
+
},
|
| 16 |
+
"norm_map": {
|
| 17 |
+
"VISUAL": "IDENTITY",
|
| 18 |
+
"STATE": "QUANTILES",
|
| 19 |
+
"ACTION": "QUANTILES"
|
| 20 |
+
}
|
| 21 |
+
},
|
| 22 |
+
"state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"registry_name": "absolute_actions_processor",
|
| 26 |
+
"config": {
|
| 27 |
+
"enabled": true
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"registry_name": "device_processor",
|
| 32 |
+
"config": {
|
| 33 |
+
"device": "cpu",
|
| 34 |
+
"float_dtype": null
|
| 35 |
+
}
|
| 36 |
+
}
|
| 37 |
+
]
|
| 38 |
+
}
|
step_004000/policy_postprocessor_step_0_unnormalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50d458bdfa5336a7d87d3addc7ed0777f0ce63e5ec74fe58d4cd5c3ecfe8e0e1
|
| 3 |
+
size 10692
|
step_004000/policy_preprocessor.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_preprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "rename_observations_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"rename_map": {
|
| 8 |
+
"observation.images.front": "observation.images.base_0_rgb",
|
| 9 |
+
"observation.images.wrist": "observation.images.left_wrist_0_rgb"
|
| 10 |
+
}
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
{
|
| 14 |
+
"registry_name": "to_batch_processor",
|
| 15 |
+
"config": {}
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"registry_name": "relative_actions_processor",
|
| 19 |
+
"config": {
|
| 20 |
+
"enabled": true,
|
| 21 |
+
"exclude_joints": [
|
| 22 |
+
"gripper"
|
| 23 |
+
],
|
| 24 |
+
"action_names": [
|
| 25 |
+
"joint_0",
|
| 26 |
+
"joint_1",
|
| 27 |
+
"joint_2",
|
| 28 |
+
"joint_3",
|
| 29 |
+
"joint_4",
|
| 30 |
+
"joint_5",
|
| 31 |
+
"joint_6",
|
| 32 |
+
"gripper"
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"registry_name": "normalizer_processor",
|
| 38 |
+
"config": {
|
| 39 |
+
"eps": 1e-08,
|
| 40 |
+
"features": {
|
| 41 |
+
"observation.images.base_0_rgb": {
|
| 42 |
+
"type": "VISUAL",
|
| 43 |
+
"shape": [
|
| 44 |
+
3,
|
| 45 |
+
224,
|
| 46 |
+
224
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
"observation.images.left_wrist_0_rgb": {
|
| 50 |
+
"type": "VISUAL",
|
| 51 |
+
"shape": [
|
| 52 |
+
3,
|
| 53 |
+
224,
|
| 54 |
+
224
|
| 55 |
+
]
|
| 56 |
+
},
|
| 57 |
+
"observation.images.right_wrist_0_rgb": {
|
| 58 |
+
"type": "VISUAL",
|
| 59 |
+
"shape": [
|
| 60 |
+
3,
|
| 61 |
+
224,
|
| 62 |
+
224
|
| 63 |
+
]
|
| 64 |
+
},
|
| 65 |
+
"observation.state": {
|
| 66 |
+
"type": "STATE",
|
| 67 |
+
"shape": [
|
| 68 |
+
8
|
| 69 |
+
]
|
| 70 |
+
},
|
| 71 |
+
"action": {
|
| 72 |
+
"type": "ACTION",
|
| 73 |
+
"shape": [
|
| 74 |
+
8
|
| 75 |
+
]
|
| 76 |
+
}
|
| 77 |
+
},
|
| 78 |
+
"norm_map": {
|
| 79 |
+
"VISUAL": "IDENTITY",
|
| 80 |
+
"STATE": "QUANTILES",
|
| 81 |
+
"ACTION": "QUANTILES"
|
| 82 |
+
}
|
| 83 |
+
},
|
| 84 |
+
"state_file": "policy_preprocessor_step_3_normalizer_processor.safetensors"
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"registry_name": "pi05_prepare_state_tokenizer_processor_step",
|
| 88 |
+
"config": {}
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"registry_name": "tokenizer_processor",
|
| 92 |
+
"config": {
|
| 93 |
+
"max_length": 200,
|
| 94 |
+
"task_key": "task",
|
| 95 |
+
"padding_side": "right",
|
| 96 |
+
"padding": "max_length",
|
| 97 |
+
"truncation": true,
|
| 98 |
+
"tokenizer_name": "google/paligemma-3b-pt-224"
|
| 99 |
+
}
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"registry_name": "device_processor",
|
| 103 |
+
"config": {
|
| 104 |
+
"device": "cuda",
|
| 105 |
+
"float_dtype": null
|
| 106 |
+
}
|
| 107 |
+
}
|
| 108 |
+
]
|
| 109 |
+
}
|
step_004000/policy_preprocessor_step_3_normalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50d458bdfa5336a7d87d3addc7ed0777f0ce63e5ec74fe58d4cd5c3ecfe8e0e1
|
| 3 |
+
size 10692
|