update readme

README.md

-# D2Go
+# <div align="center"><strong>D2Go</strong></div>
 
-D2Go is a production ready software system from FacebookResearch, which supports end-to-end model training and deployment for mobile platforms.
+## 1 简介
 
-## What's D2Go
+D2Go 是 FacebookResearch 开发的生产就绪软件系统，支持移动平台的端到端模型训练和部署。
 
-- It is a deep learning toolkit powered by [PyTorch](https://pytorch.org/) and [Detectron2](https://github.com/facebookresearch/detectron2).
-- State-of-the-art efficient backbone networks for mobile devices.
-- End-to-end model training, quantization and deployment pipeline.
-- Easy export to TorchScript format for deployment.
+## 2 安装
 
-## Installation
+组件支持组合
 
-Install PyTorch Nightly (use CUDA 10.2 as example, see details at [PyTorch Website](https://pytorch.org/get-started/)):
+| PyTorch版本 | fastpt版本 | cuBVH版本     | DTK版本  | Python版本      | 推荐编译方式 |
+| ----------- | ---------- | ------------- | -------- | --------------- | ------------ |
+| 2.5.1       | 2.1.0      | main-eeb9d5fa | >= 25.04 | 3.8、3.10、3.11 | fastpt不转码 |
+| 2.4.1       | 2.0.1      | main-eeb9d5fa | >= 25.04 | 3.8、3.10、3.11 | fastpt不转码 |
 
-```bash
-conda install pytorch torchvision cudatoolkit=10.2 -c pytorch-nightly
++ pytorch 版本大于 2.4.1 && dtk 版本大于 25.04 推荐使用 fastpt 不转码编译。
+
+### 2.1 使用pip方式安装
+
+d2go whl 包下载目录：[光和开发者社区](https://download.sourcefind.cn:65024/4/main)，选择对应的 pytorch 版本和 python 版本下载对应 d2go 的 whl 包: 
+
+```shell
+pip install torch* # 下载torch的whl包
+pip install fastpt* --no-deps # 下载fastpt的whl包
+source  /usr/local/bin/fastpt -E
+pip install d2go* # # 下载的 d2go-fastpt 的whl包
 ```
 
-Install Detectron2 (other installation options at [Detectron2](https://github.com/facebookresearch/detectron2/blob/master/INSTALL.md)):
+### 2.2 使用源码编译方式安装
 
-```bash
-python -m pip install 'git+https://github.com/facebookresearch/detectron2.git'
+#### 2.2.1 编译环境准备
+
+提供基于 fastpt 不转码编译：
+
+1. 基于光源 pytorch 基础镜像环境：镜像下载地址：[光合开发者社区](https://sourcefind.cn/#/image/dcu/pytorch)，根据 pytorch、python、dtk 及系统下载对应的镜像版本。
+2. 基于现有 python 环境：安装 pytorch，fastpt whl 包下载目录：[光合开发者社区](https://sourcefind.cn/#/image/dcu/pytorch)，根据 python、dtk 版本，下载对应 pytorch 的 whl 包。安装命令如下：
+
+```shell
+pip install torch* # 下载torch的whl包
+pip install fastpt* --no-deps # 下载fastpt的whl包, 安装顺序，先安装torch，后安装fastpt
+pip install setuptools==59.5.0 wheel pytest scikit-learn
 ```
 
-Install mobile_cv:
+#### 2.2.2 源码编译安装
 
-```bash
-python -m pip install 'git+https://github.com/facebookresearch/mobile-vision.git'
+- 代码下载
+
+```shell
+git clone http://developer.sourcefind.cn/codes/OpenDAS/d2go.git # 根据编译需要切换分支
 ```
 
-Install d2go:
+- 提供2种源码编译方式（进入 d2go 目录）：
+  - 源码编译安装：
+
+    ```bash
+    source /usr/local/bin/fastpt -C
+    python -m pip install 'git+https://github.tbedu.top/https://github.com/facebookresearch/detectron2.git'
+    python -m pip install 'git+https://github.tbedu.top/https://github.com/facebookresearch/mobile-vision.git'
+
+    python setup.py develop
+    python setup.py install
+    ```
+  - whl 包构建安装：
+
+    ```bash
+    source /usr/local/bin/fastpt -C
+
+    python setup.py bdist_wheel # 该指令用于编译whl包，执行该指令时不必执行前两个指令
+    pip install dist/d2go-0.0.1b20250528-py3-none-any.whl
+    ```
+
+#### 2.2.3 注意事项
+
++ ROCM_PATH 为 dtk 的路径，默认为 /opt/dtk；
++ 在 pytorch2.5.1 环境下编译需要支持 c++17 语法，打开setup.py文件，把文件中的 -std=c++14 修改为 -std=c++17。
+
+## 3 验证
+- 执行下面的命令测试组件：
 
 ```bash
-git clone https://github.com/facebookresearch/d2go
-cd d2go & python -m pip install .
-```
+source /usr/local/bin/fastpt -E
 
-## Get Started
+pytest -vs ./tests
 
-- [Getting Started with D2Go](./demo).
+```
 
-- See our [model zoo](./MODEL_ZOO.md) for example configs and pretrained models.
+## 4 Known Issue
 
-## License
+- 无
 
-D2Go is released under the [Apache 2.0 license](LICENSE).
+## 5 参考资料
 
+- [README_ORIGIN](README_ORIGIN.md)
+- [https://github.com/facebookresearch/d2go.git](https://github.com/facebookresearch/d2go.git)

README_ORIGIN.md

+# D2Go
+
+D2Go is a production ready software system from FacebookResearch, which supports end-to-end model training and deployment for mobile platforms.
+
+## What's D2Go
+
+- It is a deep learning toolkit powered by [PyTorch](https://pytorch.org/) and [Detectron2](https://github.com/facebookresearch/detectron2).
+- State-of-the-art efficient backbone networks for mobile devices.
+- End-to-end model training, quantization and deployment pipeline.
+- Easy export to TorchScript format for deployment.
+
+## Installation
+
+Install PyTorch Nightly (use CUDA 10.2 as example, see details at [PyTorch Website](https://pytorch.org/get-started/)):
+
+```bash
+conda install pytorch torchvision cudatoolkit=10.2 -c pytorch-nightly
+```
+
+Install Detectron2 (other installation options at [Detectron2](https://github.com/facebookresearch/detectron2/blob/master/INSTALL.md)):
+
+```bash
+python -m pip install 'git+https://github.com/facebookresearch/detectron2.git'
+```
+
+Install mobile_cv:
+
+```bash
+python -m pip install 'git+https://github.com/facebookresearch/mobile-vision.git'
+```
+
+Install d2go:
+
+```bash
+git clone https://github.com/facebookresearch/d2go
+cd d2go & python -m pip install .
+```
+
+## Get Started
+
+- [Getting Started with D2Go](./demo).
+
+- See our [model zoo](./MODEL_ZOO.md) for example configs and pretrained models.
+
+## License
+
+D2Go is released under the [Apache 2.0 license](LICENSE).
+