Skip to content

Commit 5a8e530

Browse files
MusisoulMusisoul
committed
Merge branch 'main' into train_dp
2 parents 530012c + 567edba commit 5a8e530

87 files changed

Lines changed: 6158 additions & 47 deletions

File tree

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

‎.gitignore‎

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -31,4 +31,7 @@ save_results/*
3131
*.egg-info/
3232
lightx2v_train/output_train/*
3333
lightx2v_train/output_infer/*
34+
lightx2v_ros/build
35+
lightx2v_ros/install
36+
lightx2v_ros/log
3437
.gitnexus

‎.gitmodules‎

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,3 @@
1+
[submodule "lightx2v_ros/src/simulator/simulator/libero_node/LIBERO"]
2+
path = lightx2v_ros/src/simulator/simulator/libero_node/LIBERO
3+
url = https://github.com/Lifelong-Robot-Learning/LIBERO.git

‎configs/cosmos3/cosmos3_super_i2v.json‎

Lines changed: 20 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,20 @@
1+
{
2+
"infer_steps": 50,
3+
"sample_guide_scale": 6.0,
4+
"sample_shift": 5.0,
5+
"target_height": 832,
6+
"target_width": 480,
7+
"target_video_length": 189,
8+
"target_fps": 24.0,
9+
"enable_cfg": true,
10+
"feature_caching": "NoCaching",
11+
"rms_norm_type": "one-pass",
12+
"attn_rms_norm_type": "one-pass",
13+
"rope_type": "triton",
14+
"self_attn_type": "flash_attn3",
15+
"causal_self_attn_type": "flash_attn3",
16+
"add_resolution_template": false,
17+
"add_duration_template": false,
18+
"cosmos3_meta_init": true,
19+
"vae_cpu_offload": false
20+
}

‎configs/cosmos3/cosmos3_super_t2i.json‎

Lines changed: 20 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,20 @@
1+
{
2+
"infer_steps": 50,
3+
"sample_guide_scale": 4.0,
4+
"sample_shift": 3.0,
5+
"target_height": 1024,
6+
"target_width": 1024,
7+
"target_video_length": 1,
8+
"target_fps": 24.0,
9+
"enable_cfg": true,
10+
"feature_caching": "NoCaching",
11+
"rms_norm_type": "one-pass",
12+
"attn_rms_norm_type": "one-pass",
13+
"rope_type": "triton",
14+
"self_attn_type": "flash_attn3",
15+
"causal_self_attn_type": "flash_attn3",
16+
"cosmos3_meta_init": true,
17+
"vae_cpu_offload": false,
18+
"cpu_offload": false,
19+
"offload_granularity": "block"
20+
}

‎configs/fastwam/libero_i2va.json‎

Lines changed: 15 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,15 @@
1+
{
2+
"adapter_model_path": "/data/nvme1/yongyang/nb/FastWAM/checkpoints/fastwam_release/libero_uncond_2cam224.pt",
3+
"dataset_stats_path": "/data/nvme1/yongyang/nb/FastWAM/checkpoints/fastwam_release/libero_uncond_2cam224_dataset_stats.json",
4+
"camera_size": 224,
5+
"action_chunk_size": 32,
6+
"actions_per_plan": 10,
7+
"num_steps_wait": 30,
8+
"action_infer_steps": 20,
9+
"action_sample_shift": 5.0,
10+
"action_dim_hidden": 1024,
11+
"action_dim": 7,
12+
"robot_state_dim": 8,
13+
"binarize_gripper": true,
14+
"default_prompt": "A video recorded from a robot's point of view executing the following instruction: {task_prompt}"
15+
}

‎configs/z_image/z_image_turbo_i2i.json‎

Lines changed: 9 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,9 @@
1+
{
2+
"num_channels_latents": 16,
3+
"infer_steps": 9,
4+
"attn_type": "flash_attn3",
5+
"enable_cfg": false,
6+
"sample_guide_scale": 0.0,
7+
"patch_size": 2,
8+
"i2i_denoise_strength": 1.0
9+
}

‎dockerfiles/Dockerfile_cu130_ros2_lyrical‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -37,6 +37,7 @@ RUN colcon mixin update default
3737

3838
RUN pip uninstall setuptools-scm
3939
RUN pip install catkin_pkg empy pyqt5-sip PyQt5 sip==6.15.3 setuptools==79.0.1
40+
RUN pip install hydra-core==1.2.0 robosuite==1.4.0 bddl==1.0.1 coverage==7.6.1 gym==0.25.2 mujoco==3.9.0 future easydict termcolor
4041

4142

4243
WORKDIR /root/ros2_lyrical

‎lightx2v/infer.py‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -7,6 +7,7 @@
77

88
from lightx2v.common.ops import *
99
from lightx2v.models.runners.bagel.bagel_runner import BagelRunner # noqa: F401
10+
from lightx2v.models.runners.cosmos3.cosmos3_runner import Cosmos3Runner # noqa: F401
1011
from lightx2v.models.runners.ernie_image.ernie_image_runner import ErnieImageRunner # noqa: F401
1112
from lightx2v.models.runners.flux2.flux2_runner import Flux2DevRunner, Flux2KleinRunner # noqa: F401
1213
from lightx2v.models.runners.hidream_o1_image.hidream_o1_image_runner import HidreamO1ImageRunner # noqa: F401
@@ -19,6 +20,7 @@
1920
from lightx2v.models.runners.neopp.neopp_runner import NeoppRunner # noqa: F401
2021
from lightx2v.models.runners.qwen_image.qwen_image_runner import QwenImageRunner # noqa: F401
2122
from lightx2v.models.runners.seedvr.seedvr_runner import SeedVRRunner # noqa: F401
23+
from lightx2v.models.runners.wan.fastwam_runner import FastWAMRunner # noqa: F401
2224
from lightx2v.models.runners.wan.wan_animate_runner import WanAnimateRunner # noqa: F401
2325
from lightx2v.models.runners.wan.wan_audio_runner import Wan22AudioRunner, WanAudioRunner # noqa: F401
2426
from lightx2v.models.runners.wan.wan_distill_runner import WanDistillRunner # noqa: F401
@@ -81,6 +83,7 @@ def main():
8183
"ernie_image_turbo",
8284
"hidream_o1_image",
8385
"longcat_image",
86+
"cosmos3",
8487
"wan2.2_animate",
8588
"wan2.2_s2v",
8689
"hunyuan_video_1.5",
@@ -102,6 +105,7 @@ def main():
102105
"lingbot_va",
103106
"dreamzero",
104107
"infinitetalk",
108+
"fastwam",
105109
],
106110
default="wan2.1",
107111
)

‎lightx2v/models/networks/cosmos3/__init__.py‎

Whitespace-only changes.

‎lightx2v/models/networks/cosmos3/infer/__init__.py‎

Whitespace-only changes.

0 commit comments

Comments
 (0)