You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+16-2Lines changed: 16 additions & 2 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -37,6 +37,20 @@ conda activate embodiedgen
37
37
bash install.sh basic
38
38
```
39
39
40
+
### ✅ Starting from Docker
41
+
42
+
We provide a pre-built Docker image on [Docker Hub](https://hub.docker.com/repository/docker/wangxinjie/embodiedgen) with a configured environment for your convenience. For more details, please refer to [Docker documentation](https://github.com/HorizonRobotics/EmbodiedGen/docker/README.md).
43
+
44
+
> **Note:** Model checkpoints are not included in the image, they will be automatically downloaded on first run. You still need to set up the GPT Agent manually.
45
+
46
+
```sh
47
+
IMAGE=wangxinjie/embodiedgen:env_v0.1.x
48
+
CONTAINER=EmbodiedGen-docker-${USER}
49
+
docker pull ${IMAGE}
50
+
docker run -itd --shm-size="64g" --gpus all --cap-add=SYS_PTRACE --security-opt seccomp=unconfined --privileged --net=host --name ${CONTAINER}${IMAGE}
51
+
docker exec -it ${CONTAINER} bash
52
+
```
53
+
40
54
### ✅ Setup GPT Agent
41
55
42
56
Update the API key in file: `embodied_gen/utils/gpt_config.yaml`.
@@ -52,7 +66,7 @@ You can choose between two backends for the GPT agent:
52
66
<h2id="image-to-3d">🖼️ Image-to-3D</h2>
53
67
54
68
[](https://huggingface.co/spaces/HorizonRobotics/EmbodiedGen-Image-to-3D) Generate physically plausible 3D asset URDF from single input image, offering high-quality support for digital twin systems.
55
-
69
+
(HF space is a simplified demonstration. For the full functionality, please refer to `img3d-cli`.)
56
70
<imgsrc="apps/assets/image_to_3d.jpg"alt="Image to 3D"width="900">
[](https://huggingface.co/spaces/HorizonRobotics/EmbodiedGen-Text-to-3D) Create 3D assets from text descriptions for a wide range of geometry and styles.
96
+
[](https://huggingface.co/spaces/HorizonRobotics/EmbodiedGen-Text-to-3D) Create 3D assets from text descriptions for a wide range of geometry and styles. (HF space is a simplified demonstration. For the full functionality, please refer to `text3d-cli`.)
83
97
84
98
<imgsrc="apps/assets/text_to_3d.jpg"alt="Text to 3D"width="900">
## Getting Started with Our Pre-built Docker Image
4
+
5
+
We provide pre-built Docker image on [Docker Hub](https://hub.docker.com/repository/docker/wangxinjie/embodiedgen) that includes a configured environment for your convenience.
6
+
7
+
```sh
8
+
IMAGE=wangxinjie/embodiedgen:env_v0.1.x
9
+
CONTAINER=EmbodiedGen-docker-${USER}
10
+
docker pull ${IMAGE}
11
+
docker run -itd --shm-size="64g" --gpus all --cap-add=SYS_PTRACE --security-opt seccomp=unconfined --privileged --net=host --name ${CONTAINER}${IMAGE}
12
+
docker exec -it ${CONTAINER} bash
13
+
# ref `EmbodiedGen/README.md` to get start.
14
+
```
15
+
16
+
> **Note**: Model checkpoints are not included in the default image, they will be automatically downloaded on first run. Also, you still need to configure the GPT agent manually. See the [Setup GPT Agent](https://github.com/HorizonRobotics/EmbodiedGen?tab=readme-ov-file#-setup-gpt-agent) section for detailed instructions.
17
+
18
+
If you prefer an image with all model checkpoints, you can use `wangxinjie/embodiedgen:v0.1.x`. However, please note that this image is significantly larger. We recommend using the lighter image and allowing the models to download on demand.
19
+
20
+
21
+
## Getting Started with Building from the Dockerfile
22
+
You can also build your customized docker based on our Dockerfile.
0 commit comments