(简体中文|English)
wget --no-check-certificate https://paddle-serving.bj.bcebos.com/pddet_demo/2.0/ppyolo_r50vd_dcn_1x_coco.tar
tar xf ppyolo_r50vd_dcn_1x_coco.tar
python3 -m paddle_serving_server.serve --model serving_server --port 9494 --gpu_ids 0
This model support TensorRT, if you want a faster inference, please use --use_trt.
python3 test_client.py 000000570688.jpg