Created by: wangjiawei04

为了方便Paddle Detection导出的模型可以直接用于Serving部署上线，增加了tools/export_serving_model.py 使用方式和export_model.py 一样

python tools/export_serving_model.py -c configs/faster_rcnn_r50_1x.yml  --output_dir=./inference_model -o weights=output/faster_rcnn_r50_1x/best_model

输出的结果在 inference_model/faster_rcnn_r50_1x下

infer_cfg.yml    #配置文件
serving_client  #serving模型客户端配置
serving_server #serving模型服务端配置，模型参数也在此文件夹下

启动serving 前序要求

pip install -U paddle-serving-server-gpu paddle-serving-client paddle-serving-app 

运行

GLOG_v=2 python -m paddle_serving_server_gpu.serve --model serving_server --port 9494 --gpu_id 0
python test_client.py serving_client/serving_client_conf.prototxt infer_cfg.yml 000000570688.jpg

更多资料可以参见 Faster RCNN on Paddle Serving