README.md 3.6 KB

服务端预测部署

PaddleDetection训练出来的模型可以使用Serving 部署在服务端。
本教程以在路标数据集roadsign_voc 使用configs/yolov3_mobilenet_v1_roadsign.yml算法训练的模型进行部署。
预训练模型权重文件为yolov3_mobilenet_v1_roadsign.pdparams

1. 首先验证模型

python tools/infer.py -c configs/yolov3_mobilenet_v1_roadsign.yml -o use_gpu=true weights=https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_roadsign.pdparams --infer_img=demo/road554.png

2. 安装 paddle serving

# 安装 paddle-serving-client
pip install paddle-serving-client -i https://mirror.baidu.com/pypi/simple

# 安装 paddle-serving-server
pip install paddle-serving-server -i https://mirror.baidu.com/pypi/simple

# 安装 paddle-serving-server-gpu
pip install paddle-serving-server-gpu -i https://mirror.baidu.com/pypi/simple

3. 导出模型

PaddleDetection在训练过程包括网络的前向和优化器相关参数,而在部署过程中,我们只需要前向参数,具体参考:导出模型

python tools/export_serving_model.py -c configs/yolov3_mobilenet_v1_roadsign.yml -o use_gpu=true weights=https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_roadsign.pdparams --output_dir=./inference_model

以上命令会在./inference_model文件夹下生成一个yolov3_mobilenet_v1_roadsign文件夹:

inference_model
│   ├── yolov3_mobilenet_v1_roadsign
│   │   ├── infer_cfg.yml
│   │   ├── serving_client
│   │   │   ├── serving_client_conf.prototxt
│   │   │   ├── serving_client_conf.stream.prototxt
│   │   ├── serving_server
│   │   │   ├── conv1_bn_mean
│   │   │   ├── conv1_bn_offset
│   │   │   ├── conv1_bn_scale
│   │   │   ├── ...

serving_client文件夹下serving_client_conf.prototxt详细说明了模型输入输出信息 serving_client_conf.prototxt文件内容为:

feed_var {
  name: "image"
  alias_name: "image"
  is_lod_tensor: false
  feed_type: 1
  shape: 3
  shape: 608
  shape: 608
}
feed_var {
  name: "im_size"
  alias_name: "im_size"
  is_lod_tensor: false
  feed_type: 2
  shape: 2
}
fetch_var {
  name: "multiclass_nms_0.tmp_0"
  alias_name: "multiclass_nms_0.tmp_0"
  is_lod_tensor: true
  fetch_type: 1
  shape: -1
}

4. 启动PaddleServing服务

cd inference_model/yolov3_mobilenet_v1_roadsign/

# GPU
python -m paddle_serving_server_gpu.serve --model serving_server --port 9393 --gpu_ids 0

# CPU
python -m paddle_serving_server.serve --model serving_server --port 9393

5. 测试部署的服务

准备label_list.txt文件

# 进入到导出模型文件夹
cd inference_model/yolov3_mobilenet_v1_roadsign/

# 将数据集对应的label_list.txt文件拷贝到当前文件夹下
cp ../../dataset/roadsign_voc/label_list.txt .

设置prototxt文件路径为serving_client/serving_client_conf.prototxt
设置fetchfetch=["multiclass_nms_0.tmp_0"])

测试

# 进入目录
cd inference_model/yolov3_mobilenet_v1_roadsign/

# 测试代码 test_client.py 会自动创建output文件夹,并在output下生成`bbox.json`和`road554.png`两个文件
python ../../deploy/serving/test_client.py ../../demo/road554.png