Upload Multi-Task DiT diffusion checkpoint step 100000
Browse files- checkpoint-100000/config.json +113 -0
- checkpoint-100000/model.safetensors +3 -0
- checkpoint-100000/policy_postprocessor.json +32 -0
- checkpoint-100000/policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
- checkpoint-100000/policy_preprocessor.json +83 -0
- checkpoint-100000/policy_preprocessor_step_4_normalizer_processor.safetensors +3 -0
- checkpoint-100000/train_config.json +249 -0
checkpoint-100000/config.json
ADDED
|
@@ -0,0 +1,113 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"type": "multi_task_dit",
|
| 3 |
+
"n_obs_steps": 2,
|
| 4 |
+
"input_features": {
|
| 5 |
+
"observation.state": {
|
| 6 |
+
"type": "STATE",
|
| 7 |
+
"shape": [
|
| 8 |
+
28
|
| 9 |
+
]
|
| 10 |
+
},
|
| 11 |
+
"observation.images.ego": {
|
| 12 |
+
"type": "VISUAL",
|
| 13 |
+
"shape": [
|
| 14 |
+
3,
|
| 15 |
+
240,
|
| 16 |
+
320
|
| 17 |
+
]
|
| 18 |
+
},
|
| 19 |
+
"observation.images.wrist_left": {
|
| 20 |
+
"type": "VISUAL",
|
| 21 |
+
"shape": [
|
| 22 |
+
3,
|
| 23 |
+
240,
|
| 24 |
+
320
|
| 25 |
+
]
|
| 26 |
+
},
|
| 27 |
+
"observation.images.wrist_right": {
|
| 28 |
+
"type": "VISUAL",
|
| 29 |
+
"shape": [
|
| 30 |
+
3,
|
| 31 |
+
240,
|
| 32 |
+
320
|
| 33 |
+
]
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"output_features": {
|
| 37 |
+
"action": {
|
| 38 |
+
"type": "ACTION",
|
| 39 |
+
"shape": [
|
| 40 |
+
16
|
| 41 |
+
]
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"device": "cuda",
|
| 45 |
+
"use_amp": true,
|
| 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": null,
|
| 53 |
+
"horizon": 32,
|
| 54 |
+
"n_action_steps": 24,
|
| 55 |
+
"objective": "diffusion",
|
| 56 |
+
"noise_scheduler_type": "DDIM",
|
| 57 |
+
"num_train_timesteps": 100,
|
| 58 |
+
"beta_schedule": "squaredcos_cap_v2",
|
| 59 |
+
"beta_start": 0.0001,
|
| 60 |
+
"beta_end": 0.02,
|
| 61 |
+
"prediction_type": "epsilon",
|
| 62 |
+
"clip_sample": true,
|
| 63 |
+
"clip_sample_range": 1.0,
|
| 64 |
+
"num_inference_steps": 16,
|
| 65 |
+
"sigma_min": 0.0,
|
| 66 |
+
"num_integration_steps": 100,
|
| 67 |
+
"integration_method": "euler",
|
| 68 |
+
"timestep_sampling_strategy": "beta",
|
| 69 |
+
"timestep_sampling_s": 0.999,
|
| 70 |
+
"timestep_sampling_alpha": 1.5,
|
| 71 |
+
"timestep_sampling_beta": 1.0,
|
| 72 |
+
"hidden_dim": 512,
|
| 73 |
+
"num_layers": 6,
|
| 74 |
+
"num_heads": 8,
|
| 75 |
+
"dropout": 0.1,
|
| 76 |
+
"use_positional_encoding": false,
|
| 77 |
+
"timestep_embed_dim": 256,
|
| 78 |
+
"use_rope": true,
|
| 79 |
+
"rope_base": 10000.0,
|
| 80 |
+
"vision_encoder_name": "openai/clip-vit-base-patch16",
|
| 81 |
+
"use_separate_rgb_encoder_per_camera": true,
|
| 82 |
+
"vision_encoder_lr_multiplier": 0.1,
|
| 83 |
+
"image_resize_shape": [
|
| 84 |
+
256,
|
| 85 |
+
256
|
| 86 |
+
],
|
| 87 |
+
"image_crop_shape": [
|
| 88 |
+
224,
|
| 89 |
+
224
|
| 90 |
+
],
|
| 91 |
+
"image_crop_is_random": true,
|
| 92 |
+
"text_encoder_name": "openai/clip-vit-base-patch16",
|
| 93 |
+
"tokenizer_max_length": 77,
|
| 94 |
+
"tokenizer_padding": "max_length",
|
| 95 |
+
"tokenizer_padding_side": "right",
|
| 96 |
+
"tokenizer_truncation": true,
|
| 97 |
+
"normalization_mapping": {
|
| 98 |
+
"VISUAL": "MEAN_STD",
|
| 99 |
+
"STATE": "MIN_MAX",
|
| 100 |
+
"ACTION": "MIN_MAX"
|
| 101 |
+
},
|
| 102 |
+
"optimizer_lr": 2e-05,
|
| 103 |
+
"optimizer_betas": [
|
| 104 |
+
0.95,
|
| 105 |
+
0.999
|
| 106 |
+
],
|
| 107 |
+
"optimizer_eps": 1e-08,
|
| 108 |
+
"optimizer_weight_decay": 0.0,
|
| 109 |
+
"scheduler_name": "cosine",
|
| 110 |
+
"scheduler_warmup_steps": 0,
|
| 111 |
+
"do_mask_loss_for_padding": false,
|
| 112 |
+
"drop_n_last_frames": 7
|
| 113 |
+
}
|
checkpoint-100000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a93973d8d623fa8645a2bd57c4a30bf5aee6657c34956e727e7e7791d7a8a1e0
|
| 3 |
+
size 1798487616
|
checkpoint-100000/policy_postprocessor.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
16
|
| 13 |
+
]
|
| 14 |
+
}
|
| 15 |
+
},
|
| 16 |
+
"norm_map": {
|
| 17 |
+
"VISUAL": "MEAN_STD",
|
| 18 |
+
"STATE": "MIN_MAX",
|
| 19 |
+
"ACTION": "MIN_MAX"
|
| 20 |
+
}
|
| 21 |
+
},
|
| 22 |
+
"state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"registry_name": "device_processor",
|
| 26 |
+
"config": {
|
| 27 |
+
"device": "cpu",
|
| 28 |
+
"float_dtype": null
|
| 29 |
+
}
|
| 30 |
+
}
|
| 31 |
+
]
|
| 32 |
+
}
|
checkpoint-100000/policy_postprocessor_step_0_unnormalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ab3e214d16695c7b5055a1cc72adbb26bc63b4a912b89c433691e2ba098fb24
|
| 3 |
+
size 9872
|
checkpoint-100000/policy_preprocessor.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "policy_preprocessor",
|
| 3 |
+
"steps": [
|
| 4 |
+
{
|
| 5 |
+
"registry_name": "rename_observations_processor",
|
| 6 |
+
"config": {
|
| 7 |
+
"rename_map": {}
|
| 8 |
+
}
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"registry_name": "to_batch_processor",
|
| 12 |
+
"config": {}
|
| 13 |
+
},
|
| 14 |
+
{
|
| 15 |
+
"registry_name": "tokenizer_processor",
|
| 16 |
+
"config": {
|
| 17 |
+
"max_length": 77,
|
| 18 |
+
"task_key": "task",
|
| 19 |
+
"padding_side": "right",
|
| 20 |
+
"padding": "max_length",
|
| 21 |
+
"truncation": true,
|
| 22 |
+
"tokenizer_name": "openai/clip-vit-base-patch16"
|
| 23 |
+
}
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"registry_name": "device_processor",
|
| 27 |
+
"config": {
|
| 28 |
+
"device": "cuda",
|
| 29 |
+
"float_dtype": null
|
| 30 |
+
}
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"registry_name": "normalizer_processor",
|
| 34 |
+
"config": {
|
| 35 |
+
"eps": 1e-08,
|
| 36 |
+
"features": {
|
| 37 |
+
"observation.state": {
|
| 38 |
+
"type": "STATE",
|
| 39 |
+
"shape": [
|
| 40 |
+
28
|
| 41 |
+
]
|
| 42 |
+
},
|
| 43 |
+
"observation.images.ego": {
|
| 44 |
+
"type": "VISUAL",
|
| 45 |
+
"shape": [
|
| 46 |
+
3,
|
| 47 |
+
240,
|
| 48 |
+
320
|
| 49 |
+
]
|
| 50 |
+
},
|
| 51 |
+
"observation.images.wrist_left": {
|
| 52 |
+
"type": "VISUAL",
|
| 53 |
+
"shape": [
|
| 54 |
+
3,
|
| 55 |
+
240,
|
| 56 |
+
320
|
| 57 |
+
]
|
| 58 |
+
},
|
| 59 |
+
"observation.images.wrist_right": {
|
| 60 |
+
"type": "VISUAL",
|
| 61 |
+
"shape": [
|
| 62 |
+
3,
|
| 63 |
+
240,
|
| 64 |
+
320
|
| 65 |
+
]
|
| 66 |
+
},
|
| 67 |
+
"action": {
|
| 68 |
+
"type": "ACTION",
|
| 69 |
+
"shape": [
|
| 70 |
+
16
|
| 71 |
+
]
|
| 72 |
+
}
|
| 73 |
+
},
|
| 74 |
+
"norm_map": {
|
| 75 |
+
"VISUAL": "MEAN_STD",
|
| 76 |
+
"STATE": "MIN_MAX",
|
| 77 |
+
"ACTION": "MIN_MAX"
|
| 78 |
+
}
|
| 79 |
+
},
|
| 80 |
+
"state_file": "policy_preprocessor_step_4_normalizer_processor.safetensors"
|
| 81 |
+
}
|
| 82 |
+
]
|
| 83 |
+
}
|
checkpoint-100000/policy_preprocessor_step_4_normalizer_processor.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ab3e214d16695c7b5055a1cc72adbb26bc63b4a912b89c433691e2ba098fb24
|
| 3 |
+
size 9872
|
checkpoint-100000/train_config.json
ADDED
|
@@ -0,0 +1,249 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"dataset": {
|
| 3 |
+
"repo_id": "rm65b-sort-v0",
|
| 4 |
+
"root": "/scratch3/workspace/junyicao_umass_edu-physgen-v9/ckpt/vla_training/rm65b-sort-v0",
|
| 5 |
+
"episodes": null,
|
| 6 |
+
"image_transforms": {
|
| 7 |
+
"enable": true,
|
| 8 |
+
"max_num_transforms": 3,
|
| 9 |
+
"random_order": false,
|
| 10 |
+
"tfs": {
|
| 11 |
+
"brightness": {
|
| 12 |
+
"weight": 1.0,
|
| 13 |
+
"type": "ColorJitter",
|
| 14 |
+
"kwargs": {
|
| 15 |
+
"brightness": [
|
| 16 |
+
0.8,
|
| 17 |
+
1.2
|
| 18 |
+
]
|
| 19 |
+
}
|
| 20 |
+
},
|
| 21 |
+
"contrast": {
|
| 22 |
+
"weight": 1.0,
|
| 23 |
+
"type": "ColorJitter",
|
| 24 |
+
"kwargs": {
|
| 25 |
+
"contrast": [
|
| 26 |
+
0.8,
|
| 27 |
+
1.2
|
| 28 |
+
]
|
| 29 |
+
}
|
| 30 |
+
},
|
| 31 |
+
"saturation": {
|
| 32 |
+
"weight": 1.0,
|
| 33 |
+
"type": "ColorJitter",
|
| 34 |
+
"kwargs": {
|
| 35 |
+
"saturation": [
|
| 36 |
+
0.5,
|
| 37 |
+
1.5
|
| 38 |
+
]
|
| 39 |
+
}
|
| 40 |
+
},
|
| 41 |
+
"hue": {
|
| 42 |
+
"weight": 1.0,
|
| 43 |
+
"type": "ColorJitter",
|
| 44 |
+
"kwargs": {
|
| 45 |
+
"hue": [
|
| 46 |
+
-0.05,
|
| 47 |
+
0.05
|
| 48 |
+
]
|
| 49 |
+
}
|
| 50 |
+
},
|
| 51 |
+
"sharpness": {
|
| 52 |
+
"weight": 1.0,
|
| 53 |
+
"type": "SharpnessJitter",
|
| 54 |
+
"kwargs": {
|
| 55 |
+
"sharpness": [
|
| 56 |
+
0.5,
|
| 57 |
+
1.5
|
| 58 |
+
]
|
| 59 |
+
}
|
| 60 |
+
},
|
| 61 |
+
"affine": {
|
| 62 |
+
"weight": 1.0,
|
| 63 |
+
"type": "RandomAffine",
|
| 64 |
+
"kwargs": {
|
| 65 |
+
"degrees": [
|
| 66 |
+
-5.0,
|
| 67 |
+
5.0
|
| 68 |
+
],
|
| 69 |
+
"translate": [
|
| 70 |
+
0.05,
|
| 71 |
+
0.05
|
| 72 |
+
]
|
| 73 |
+
}
|
| 74 |
+
}
|
| 75 |
+
}
|
| 76 |
+
},
|
| 77 |
+
"revision": null,
|
| 78 |
+
"use_imagenet_stats": true,
|
| 79 |
+
"video_backend": "torchcodec",
|
| 80 |
+
"streaming": false
|
| 81 |
+
},
|
| 82 |
+
"env": null,
|
| 83 |
+
"policy": {
|
| 84 |
+
"type": "multi_task_dit",
|
| 85 |
+
"n_obs_steps": 2,
|
| 86 |
+
"input_features": {
|
| 87 |
+
"observation.state": {
|
| 88 |
+
"type": "STATE",
|
| 89 |
+
"shape": [
|
| 90 |
+
28
|
| 91 |
+
]
|
| 92 |
+
},
|
| 93 |
+
"observation.images.ego": {
|
| 94 |
+
"type": "VISUAL",
|
| 95 |
+
"shape": [
|
| 96 |
+
3,
|
| 97 |
+
240,
|
| 98 |
+
320
|
| 99 |
+
]
|
| 100 |
+
},
|
| 101 |
+
"observation.images.wrist_left": {
|
| 102 |
+
"type": "VISUAL",
|
| 103 |
+
"shape": [
|
| 104 |
+
3,
|
| 105 |
+
240,
|
| 106 |
+
320
|
| 107 |
+
]
|
| 108 |
+
},
|
| 109 |
+
"observation.images.wrist_right": {
|
| 110 |
+
"type": "VISUAL",
|
| 111 |
+
"shape": [
|
| 112 |
+
3,
|
| 113 |
+
240,
|
| 114 |
+
320
|
| 115 |
+
]
|
| 116 |
+
}
|
| 117 |
+
},
|
| 118 |
+
"output_features": {
|
| 119 |
+
"action": {
|
| 120 |
+
"type": "ACTION",
|
| 121 |
+
"shape": [
|
| 122 |
+
16
|
| 123 |
+
]
|
| 124 |
+
}
|
| 125 |
+
},
|
| 126 |
+
"device": "cuda",
|
| 127 |
+
"use_amp": true,
|
| 128 |
+
"use_peft": false,
|
| 129 |
+
"push_to_hub": false,
|
| 130 |
+
"repo_id": null,
|
| 131 |
+
"private": null,
|
| 132 |
+
"tags": null,
|
| 133 |
+
"license": null,
|
| 134 |
+
"pretrained_path": null,
|
| 135 |
+
"horizon": 32,
|
| 136 |
+
"n_action_steps": 24,
|
| 137 |
+
"objective": "diffusion",
|
| 138 |
+
"noise_scheduler_type": "DDIM",
|
| 139 |
+
"num_train_timesteps": 100,
|
| 140 |
+
"beta_schedule": "squaredcos_cap_v2",
|
| 141 |
+
"beta_start": 0.0001,
|
| 142 |
+
"beta_end": 0.02,
|
| 143 |
+
"prediction_type": "epsilon",
|
| 144 |
+
"clip_sample": true,
|
| 145 |
+
"clip_sample_range": 1.0,
|
| 146 |
+
"num_inference_steps": 16,
|
| 147 |
+
"sigma_min": 0.0,
|
| 148 |
+
"num_integration_steps": 100,
|
| 149 |
+
"integration_method": "euler",
|
| 150 |
+
"timestep_sampling_strategy": "beta",
|
| 151 |
+
"timestep_sampling_s": 0.999,
|
| 152 |
+
"timestep_sampling_alpha": 1.5,
|
| 153 |
+
"timestep_sampling_beta": 1.0,
|
| 154 |
+
"hidden_dim": 512,
|
| 155 |
+
"num_layers": 6,
|
| 156 |
+
"num_heads": 8,
|
| 157 |
+
"dropout": 0.1,
|
| 158 |
+
"use_positional_encoding": false,
|
| 159 |
+
"timestep_embed_dim": 256,
|
| 160 |
+
"use_rope": true,
|
| 161 |
+
"rope_base": 10000.0,
|
| 162 |
+
"vision_encoder_name": "openai/clip-vit-base-patch16",
|
| 163 |
+
"use_separate_rgb_encoder_per_camera": true,
|
| 164 |
+
"vision_encoder_lr_multiplier": 0.1,
|
| 165 |
+
"image_resize_shape": [
|
| 166 |
+
256,
|
| 167 |
+
256
|
| 168 |
+
],
|
| 169 |
+
"image_crop_shape": [
|
| 170 |
+
224,
|
| 171 |
+
224
|
| 172 |
+
],
|
| 173 |
+
"image_crop_is_random": true,
|
| 174 |
+
"text_encoder_name": "openai/clip-vit-base-patch16",
|
| 175 |
+
"tokenizer_max_length": 77,
|
| 176 |
+
"tokenizer_padding": "max_length",
|
| 177 |
+
"tokenizer_padding_side": "right",
|
| 178 |
+
"tokenizer_truncation": true,
|
| 179 |
+
"normalization_mapping": {
|
| 180 |
+
"VISUAL": "MEAN_STD",
|
| 181 |
+
"STATE": "MIN_MAX",
|
| 182 |
+
"ACTION": "MIN_MAX"
|
| 183 |
+
},
|
| 184 |
+
"optimizer_lr": 2e-05,
|
| 185 |
+
"optimizer_betas": [
|
| 186 |
+
0.95,
|
| 187 |
+
0.999
|
| 188 |
+
],
|
| 189 |
+
"optimizer_eps": 1e-08,
|
| 190 |
+
"optimizer_weight_decay": 0.0,
|
| 191 |
+
"scheduler_name": "cosine",
|
| 192 |
+
"scheduler_warmup_steps": 0,
|
| 193 |
+
"do_mask_loss_for_padding": false,
|
| 194 |
+
"drop_n_last_frames": 7
|
| 195 |
+
},
|
| 196 |
+
"output_dir": "/scratch3/workspace/junyicao_umass_edu-physgen-v9/output/vla_training/dit_diffusion_rm65b_sort_v0",
|
| 197 |
+
"job_name": "multi_task_dit",
|
| 198 |
+
"resume": false,
|
| 199 |
+
"seed": 42,
|
| 200 |
+
"cudnn_deterministic": false,
|
| 201 |
+
"num_workers": 8,
|
| 202 |
+
"batch_size": 32,
|
| 203 |
+
"steps": 100000,
|
| 204 |
+
"eval_freq": 5000,
|
| 205 |
+
"log_freq": 100,
|
| 206 |
+
"tolerance_s": 0.0001,
|
| 207 |
+
"save_checkpoint": true,
|
| 208 |
+
"save_freq": 5000,
|
| 209 |
+
"use_policy_training_preset": true,
|
| 210 |
+
"optimizer": {
|
| 211 |
+
"type": "adam",
|
| 212 |
+
"lr": 2e-05,
|
| 213 |
+
"weight_decay": 0.0,
|
| 214 |
+
"grad_clip_norm": 10.0,
|
| 215 |
+
"betas": [
|
| 216 |
+
0.95,
|
| 217 |
+
0.999
|
| 218 |
+
],
|
| 219 |
+
"eps": 1e-08
|
| 220 |
+
},
|
| 221 |
+
"scheduler": {
|
| 222 |
+
"type": "diffuser",
|
| 223 |
+
"num_warmup_steps": 0,
|
| 224 |
+
"name": "cosine"
|
| 225 |
+
},
|
| 226 |
+
"eval": {
|
| 227 |
+
"n_episodes": 50,
|
| 228 |
+
"batch_size": 50,
|
| 229 |
+
"use_async_envs": false
|
| 230 |
+
},
|
| 231 |
+
"wandb": {
|
| 232 |
+
"enable": false,
|
| 233 |
+
"disable_artifact": true,
|
| 234 |
+
"project": "MultiTaskDiT",
|
| 235 |
+
"entity": null,
|
| 236 |
+
"notes": null,
|
| 237 |
+
"run_id": null,
|
| 238 |
+
"mode": null,
|
| 239 |
+
"add_tags": true
|
| 240 |
+
},
|
| 241 |
+
"peft": null,
|
| 242 |
+
"use_rabc": false,
|
| 243 |
+
"rabc_progress_path": null,
|
| 244 |
+
"rabc_kappa": 0.01,
|
| 245 |
+
"rabc_epsilon": 1e-06,
|
| 246 |
+
"rabc_head_mode": "sparse",
|
| 247 |
+
"rename_map": {},
|
| 248 |
+
"checkpoint_path": null
|
| 249 |
+
}
|