KoeYe commited on
Commit
5e3c8fb
·
verified ·
1 Parent(s): 226ef5c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +19 -16
README.md CHANGED
@@ -52,7 +52,7 @@ SimWorld is built on Unreal Engine 5 and offers core capabilities to meet the ne
52
  **SimWorld** consists of three layers:
53
  - the Unreal Engine Backend, providing diverse and open-ended environments, rich assets and realistic physics simulation;
54
  - the Environment layer, supporting procedural city generation, language-driven scene editing, gym-like APIs for LLM/VLM agents and traffic simulation;
55
- - the Agent layer, enabling LLM/VLM agents to reason over multimodal observations and history while executing actions via a local action planner;
56
 
57
  SimWorld's architecture is designed to be modular and flexible, supporting an array of functionalities such as dynamic world generation, agent control, and performance benchmarking. The components are seamlessly integrated to provide a robust platform for **Embodied AI** and **Agents** research and applications.
58
 
@@ -69,9 +69,11 @@ simworld/ # Python package
69
  map/ # Basic map class and waypoint system
70
  traffic/ # Traffic system
71
  utils/ # Utility functions
 
 
72
  data/ # Necessary input data
73
  config/ # Example configuration file and user configuration file
74
- scripts/ # Examples of usage, such as layout generation and traffic simulation
75
  docs/ # Documentation source files
76
  README.md
77
  ```
@@ -89,17 +91,16 @@ pip install -e .
89
  ```
90
 
91
  + UE server
92
- Download the SimWorld server executable from S3:
93
 
94
  We offer two versions of the SimWorld UE package: the base version, which comes with an empty map, and the additional environments version, which provides extra pre-defined environments for more diverse simulation scenarios. Both versions include all the core features of SimWorld.
95
 
96
  | Platform | Package | Scenes/Maps Included | Download | Notes |
97
  | --- | --- | --- | --- | --- |
98
- | Windows | Base | Empty map for procedural generation | [Download (Base)](https://simworld-release.s3.us-east-1.amazonaws.com/SimWorld-Win64-v0_1_0-Foundation.zip) | Full agent features; smaller download. |
99
- | Windows | Additional Environments | 100+ maps (including the empty one) | [Download (100+ Maps)](https://simworld-release.s3.us-east-1.amazonaws.com/SimWorld-Win64-v0_1_0-100Maps.zip) | Full agent features; larger download. |
100
- | Linux | Base | Empty map for procedural generation | [Download (Base)](https://simworld-release.s3.us-east-1.amazonaws.com/SimWorld-Linux-v0_1_0-Foundation.zip) | Full agent features; smaller download. |
101
- | Linux | Additional Environments | 100+ maps (including the empty one) | [Download (100+ Maps)](https://simworld-release.s3.us-east-1.amazonaws.com/SimWorld-Linux-v0_1_0-100Maps.zip) | Full agent features; larger download. |
102
 
 
103
 
104
  **Note:**
105
  1. Please check the [documentation](https://simworld.readthedocs.io/en/latest/getting_started/additional_environments.html#usage) for usage instructions of the **100+ Maps** version.
@@ -116,7 +117,7 @@ SimWorld uses YAML-formatted configuration files for system settings. The defaul
116
  - `simworld/config/default.yaml` serves as the default configuration file.
117
  - `config/example.yaml` is provided as a template for custom configurations.
118
 
119
- Users can switch between different configurations by specifying a custom configuration file path through the `Config` class:
120
 
121
  To set up your own configuration:
122
 
@@ -125,7 +126,7 @@ To set up your own configuration:
125
  cp config/example.yaml config/your_config.yaml
126
  ```
127
 
128
- 2. Modify the configuration values in `your_config.yaml` according to your needs
129
 
130
  3. Load your custom configuration in your code:
131
  ```python
@@ -161,11 +162,6 @@ from simworld.llm.base_llm import BaseLLM
161
  from simworld.local_planner.local_planner import LocalPlanner
162
  from simworld.llm.a2a_llm import A2ALLM
163
 
164
-
165
- # Connect to the running Unreal Engine instance via UnrealCV
166
- ucv = UnrealCV()
167
- comm = Communicator(ucv)
168
-
169
  class Agent:
170
  def __init__(self, goal):
171
  self.goal = goal
@@ -228,8 +224,11 @@ class Environment:
228
 
229
  return observation, reward
230
 
231
-
232
  if __name__ == "__main__":
 
 
 
 
233
  # Create the environment wrapper
234
  agent = Agent(goal='Go to (1700, -1700) and pick up GEN_BP_Box_1_C.')
235
  env = Environment(comm)
@@ -242,4 +241,8 @@ if __name__ == "__main__":
242
  obs, reward = env.step(action)
243
  print(f"obs: {obs}, reward: {reward}")
244
  # Plug this into your RL loop / logging as needed
245
- ```
 
 
 
 
 
52
  **SimWorld** consists of three layers:
53
  - the Unreal Engine Backend, providing diverse and open-ended environments, rich assets and realistic physics simulation;
54
  - the Environment layer, supporting procedural city generation, language-driven scene editing, gym-like APIs for LLM/VLM agents and traffic simulation;
55
+ - the Agent layer, enabling LLM/VLM agents to reason over multimodal observations and history while executing actions via a local action planner.
56
 
57
  SimWorld's architecture is designed to be modular and flexible, supporting an array of functionalities such as dynamic world generation, agent control, and performance benchmarking. The components are seamlessly integrated to provide a robust platform for **Embodied AI** and **Agents** research and applications.
58
 
 
69
  map/ # Basic map class and waypoint system
70
  traffic/ # Traffic system
71
  utils/ # Utility functions
72
+ data/ # Default data files, e.g., object categories
73
+ weather/ # Weather system
74
  data/ # Necessary input data
75
  config/ # Example configuration file and user configuration file
76
+ examples/ # Examples of usage, such as layout generation and traffic simulation
77
  docs/ # Documentation source files
78
  README.md
79
  ```
 
91
  ```
92
 
93
  + UE server
94
+ Download the SimWorld server executable from huggingface. Choose the version according to your OS and the edition you want to use.
95
 
96
  We offer two versions of the SimWorld UE package: the base version, which comes with an empty map, and the additional environments version, which provides extra pre-defined environments for more diverse simulation scenarios. Both versions include all the core features of SimWorld.
97
 
98
  | Platform | Package | Scenes/Maps Included | Download | Notes |
99
  | --- | --- | --- | --- | --- |
100
+ | Windows | Base | Empty map for procedural generation | [Download](https://huggingface.co/datasets/SimWorld-AI/SimWorld/resolve/main/Base/Windows.zip) | Full agent features; smaller download. |
101
+ | Linux | Base | Empty map for procedural generation | [Download](https://huggingface.co/datasets/SimWorld-AI/SimWorld/resolve/main/Base/Linux.zip) | Full agent features; smaller download. |
 
 
102
 
103
+ Additional environment paks are available on the [environments paks page](https://huggingface.co/datasets/SimWorld-AI/SimWorld/tree/main/AdditionEnvironmentPaks). You may download them as needed according to the OS you are using.
104
 
105
  **Note:**
106
  1. Please check the [documentation](https://simworld.readthedocs.io/en/latest/getting_started/additional_environments.html#usage) for usage instructions of the **100+ Maps** version.
 
117
  - `simworld/config/default.yaml` serves as the default configuration file.
118
  - `config/example.yaml` is provided as a template for custom configurations.
119
 
120
+ Users can switch between different configurations by specifying a custom configuration file path through the `Config` class.
121
 
122
  To set up your own configuration:
123
 
 
126
  cp config/example.yaml config/your_config.yaml
127
  ```
128
 
129
+ 2. Modify the configuration values in `your_config.yaml` according to your needs.
130
 
131
  3. Load your custom configuration in your code:
132
  ```python
 
162
  from simworld.local_planner.local_planner import LocalPlanner
163
  from simworld.llm.a2a_llm import A2ALLM
164
 
 
 
 
 
 
165
  class Agent:
166
  def __init__(self, goal):
167
  self.goal = goal
 
224
 
225
  return observation, reward
226
 
 
227
  if __name__ == "__main__":
228
+ # Connect to the running Unreal Engine instance via UnrealCV
229
+ ucv = UnrealCV()
230
+ comm = Communicator(ucv)
231
+
232
  # Create the environment wrapper
233
  agent = Agent(goal='Go to (1700, -1700) and pick up GEN_BP_Box_1_C.')
234
  env = Environment(comm)
 
241
  obs, reward = env.step(action)
242
  print(f"obs: {obs}, reward: {reward}")
243
  # Plug this into your RL loop / logging as needed
244
+ ```
245
+
246
+ ## Star History
247
+
248
+ [![Star History Chart](https://api.star-history.com/svg?repos=SimWorld-AI/SimWorld&type=date&legend=bottom-right)](https://www.star-history.com/#SimWorld-AI/SimWorld&type=date&legend=bottom-right)