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
`MotrixLab`is a reinforcement learning framework based on the [MotrixSim](https://github.com/Motphys/motrixsim-docs)simulation engine, designed specifically for robot simulation and training. This project provides a complete reinforcement learning development platform that integrates multiple simulation environments and training frameworks.
-**motrix_envs**: Various RL simulation environments built on MotrixSim, defining observation, action, and reward. Framework-agnostic and currently supports MotrixSim's CPU backend
15
+
-**motrix_rl**: Integrates RL frameworks and uses various environment parameters from motrix_envs for training. Currently supports SKRL framework's PPO algorithm
14
16
15
-
> 文档地址：https://motrixlab.readthedocs.io
17
+
> Documentation: https://motrixlab.readthedocs.io
16
18
17
-
## 主要特性
19
+
## Key Features
18
20
19
-
-**统一接口**: 提供简洁统一的强化学习训练和评估接口
20
-
-**多后端支持**: 支持 JAX 和 PyTorch 训练后端，可根据硬件环境灵活选择
21
-
-**丰富环境**: 包含基础控制、运动、操作等多种机器人仿真环境
22
-
-**高性能仿真**: 基于 MotrixSim 的高性能物理仿真引擎
23
-
-**可视化训练**: 支持实时渲染和训练过程可视化
21
+
-**Unified Interface**: Provides a concise and unified reinforcement learning training and evaluation interface
22
+
-**Multi-backend Support**: Supports JAX and PyTorch training backends, with flexible selection based on hardware environment
23
+
-**Rich Environments**: Includes various robot simulation environments such as basic control, locomotion, and manipulation tasks
24
+
-**High-performance Simulation**: Built on MotrixSim's high-performance physics simulation engine
25
+
-**Visual Training**: Supports real-time rendering and training process visualization
SKRL framework supports JAX(Flax) or PyTorch as training backends. You can also choose to install only one training backend based on your hardware environment:
50
52
51
-
安装 JAX 作为训练后端（仅支持 Linux 平台）：
53
+
Install JAX as training backend (Linux only):
52
54
53
55
```bash
54
56
uv sync --all-packages --extra skrl-jax
55
57
```
56
58
57
-
安装 PyTorch 作为训练后端：
59
+
Install PyTorch as training backend:
58
60
59
61
```bash
60
62
uv sync --all-packages --extra skrl-torch
61
63
```
62
64
63
-
## 🎯 使用指南
65
+
## 🎯 Usage Guide
64
66
65
-
### 环境可视化
67
+
### Environment Visualization
66
68
67
-
查看环境而不执行训练：
69
+
View environments without executing training:
68
70
69
71
```bash
70
72
uv run scripts/view.py --env cartpole
71
73
```
72
74
73
-
### 训练模型
75
+
### Model Training
74
76
75
77
```bash
76
78
uv run scripts/train.py --env cartpole
77
79
```
78
80
79
-
训练结果会保存在 `runs/{env-name}/`目录下。
81
+
Training results are saved in the `runs/{env-name}/`directory.
80
82
81
-
通过 TensorBoard 查看训练数据：
83
+
View training data through TensorBoard:
82
84
83
85
```bash
84
86
uv run tensorboard --logdir runs/{env-name}
85
87
```
86
88
87
-
### 模型推理
89
+
### Model Inference
88
90
89
-
```
91
+
```bash
90
92
uv run scripts/play.py --env cartpole
91
93
```
92
94
93
-
更多使用方式请参考[用户文档](https://motrixlab.readthedocs.io)
95
+
For more usage methods, please refer to the [User Documentation](https://motrixlab.readthedocs.io)
94
96
95
-
## 📬 联系方式
97
+
## 📬 Contact
96
98
97
-
有问题或建议？欢迎通过以下方式联系我们：
99
+
Have questions or suggestions? Feel free to contact us through:
0 commit comments