Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion README.md
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@

- [飞桨自定义接入硬件后端(昇腾NPU)](backends/npu/README_cn.md)
- [飞桨自定义接入硬件后端(寒武纪MLU)](backends/mlu/README_cn.md)
- [飞桨自定义接入硬件后端(英特尔GPU)](backends/intel_gpu/README.md)
- [飞桨自定义接入硬件后端(英特尔GPU)](backends/intel_gpu/README_cn.md)
- [飞桨自定义接入硬件后端(苹果MPS)](backends/mps/README.md)
- [飞桨自定义接入硬件后端(壁仞GPU)](backends/biren_gpu/README_cn.md)
- [飞桨自定义接入硬件后端(燧原GCU)](backends/gcu/README_cn.md)
Expand Down
4 changes: 3 additions & 1 deletion backends/intel_gpu/README.md
Original file line number Diff line number Diff line change
@@ -1,8 +1,10 @@
# PaddlePaddle Custom Device Implementaion for Custom Intel GPU

English | [简体中文](./README_cn.md)

Please refer to the following steps to compile, install and verify the custom device implementaion for Custom Intel GPU.

## Activate oneapi env vars
## Activate Intel oneAPI env vars

```bash
source load.sh
Expand Down
53 changes: 53 additions & 0 deletions backends/intel_gpu/README_cn.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,53 @@
# 飞桨自定义接入硬件后端(Intel GPU)

简体中文 | [English](./README.md)

请参考以下步骤进行硬件后端(Intel GPU)的编译安装与验证

## 激活 Intel oneAPI 环境变量

```bash
source load.sh
```

## 获取源码

```bash
# 克隆代码
git clone --recursive https://github.com/PaddlePaddle/PaddleCustomDevice
cd PaddleCustomDevice

# 请执行以下命令,以保证checkout最新的Paddle源码
git submodule sync
git submodule update --remote --init --recursive
```

## 编译安装

```bash
# 进入 Intel GPU Backend 目录
cd backends/intel_gpu

# 编译前,请确保已安装 Paddle,您可以运行以下命令
pip install paddlepaddle==0.0.0 -f https://www.paddlepaddle.org.cn/whl/linux/cpu-mkl/develop.html

# 创建编译目录并编译
mkdir build && cd build

cmake ..
make -j $(nproc)

# 编译产出在dist路径下,使用pip安装
pip install dist/paddle_custom_intel_gpu*.whl
```

## 验证

```bash
# 检查插件状态
python -c "import paddle; print('intel_gpu' in paddle.device.get_all_custom_device_type())"

# 预期输出
True

```
4 changes: 2 additions & 2 deletions backends/intel_hpu/README_cn.md
Original file line number Diff line number Diff line change
Expand Up @@ -40,12 +40,12 @@ pip install dist/paddle_habana_hpu*.whl
```bash
# 列出可用硬件后端
python -c "import paddle; print(paddle.device.get_all_custom_device_type())"
# 期待输出以下结果
# 预期输出以下结果
['intel_hpu']

# 运行简单模型
python ../tests/test_MNIST_model.py
# 期待输出以下类似结果
# 预期输出以下类似结果
... ...
Epoch 0 step 0, Loss = [2.2956038], Accuracy = 0.15625
Epoch 0 step 100, Loss = [2.1552896], Accuracy = 0.3125
Expand Down