diff --git a/deploy/slim/quant_post_static.py b/deploy/slim/quant_post_static.py index 5c8469794ad29e18dad15f985b611e423fd4b474..20507c66ad1ed583c2baf1bae7e812e0364e015e 100644 --- a/deploy/slim/quant_post_static.py +++ b/deploy/slim/quant_post_static.py @@ -43,6 +43,7 @@ def main(): 'inference.pdiparams')) config["DataLoader"]["Eval"]["sampler"]["batch_size"] = 1 config["DataLoader"]["Eval"]["loader"]["num_workers"] = 0 + init_logger() device = paddle.set_device("cpu") train_dataloader = build_dataloader(config["DataLoader"], "Eval", device, @@ -67,6 +68,7 @@ def main(): quantize_model_path=os.path.join( config["Global"]["save_inference_dir"], "quant_post_static_model"), sample_generator=sample_generator(train_dataloader), + batch_size=config["DataLoader"]["Eval"]["sampler"]["batch_size"], batch_nums=10) diff --git a/test_tipc/config/MobileNetV3/MobileNetV3_large_x1_0-KL_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/config/MobileNetV3/MobileNetV3_large_x1_0-KL_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt new file mode 100644 index 0000000000000000000000000000000000000000..6393d49a2f4b6d0e7228e88f046e738230d0c542 --- /dev/null +++ b/test_tipc/config/MobileNetV3/MobileNetV3_large_x1_0-KL_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt @@ -0,0 +1,18 @@ +===========================cpp_infer_params=========================== +model_name:MobileNetV3_large_x1_0_KL +cpp_infer_type:cls +cls_inference_model_dir:./MobileNetV3_large_x1_0_kl_quant_infer/ +det_inference_model_dir: +cls_inference_url:https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/inference/slim_model/MobileNetV3_large_x1_0_kl_quant_infer.tar +det_inference_url: +infer_quant:False +inference_cmd:./deploy/cpp/build/clas_system -c inference_cls.yaml +use_gpu:True|False +enable_mkldnn:False +cpu_threads:1 +batch_size:1 +use_tensorrt:False +precision:fp32 +image_dir:./dataset/ILSVRC2012/val/ILSVRC2012_val_00000001.JPEG +benchmark:False +generate_yaml_cmd:python3.7 test_tipc/generate_cpp_yaml.py diff --git a/test_tipc/config/MobileNetV3/MobileNetV3_large_x1_0-KL_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/config/MobileNetV3/MobileNetV3_large_x1_0-KL_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt new file mode 100644 index 0000000000000000000000000000000000000000..05efbacc76cbfd69279d658350abca9cfc607bd5 --- /dev/null +++ b/test_tipc/config/MobileNetV3/MobileNetV3_large_x1_0-KL_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt @@ -0,0 +1,14 @@ +===========================serving_params=========================== +model_name:MobileNetV3_large_x1_0_KL +python:python3.7 +inference_model_url:https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/inference/slim_model/MobileNetV3_large_x1_0_kl_quant_infer.tar +trans_model:-m paddle_serving_client.convert +--dirname:./deploy/paddleserving/MobileNetV3_large_x1_0_kl_quant_infer/ +--model_filename:inference.pdmodel +--params_filename:inference.pdiparams +--serving_server:./deploy/paddleserving/MobileNetV3_large_x1_0_kl_quant_serving/ +--serving_client:./deploy/paddleserving/MobileNetV3_large_x1_0_kl_quant_client/ +serving_dir:./deploy/paddleserving +web_service:null +--use_gpu:0|null +pipline:test_cpp_serving_client.py diff --git a/test_tipc/config/MobileNetV3/MobileNetV3_large_x1_0-KL_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/config/MobileNetV3/MobileNetV3_large_x1_0-KL_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt new file mode 100644 index 0000000000000000000000000000000000000000..32b506afd594b486d6fe759ee2d1556f9b049a8c --- /dev/null +++ b/test_tipc/config/MobileNetV3/MobileNetV3_large_x1_0-KL_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt @@ -0,0 +1,14 @@ +===========================serving_params=========================== +model_name:MobileNetV3_large_x1_0_KL +python:python3.7 +inference_model_url:https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/inference/slim_model/MobileNetV3_large_x1_0_kl_quant_infer.tar +trans_model:-m paddle_serving_client.convert +--dirname:./deploy/paddleserving/MobileNetV3_large_x1_0_kl_quant_infer/ +--model_filename:inference.pdmodel +--params_filename:inference.pdiparams +--serving_server:./deploy/paddleserving/MobileNetV3_large_x1_0_kl_quant_serving/ +--serving_client:./deploy/paddleserving/MobileNetV3_large_x1_0_kl_quant_client/ +serving_dir:./deploy/paddleserving +web_service:classification_web_service.py +--use_gpu:0|null +pipline:pipeline_http_client.py diff --git a/test_tipc/config/ResNet/ResNet50_vd-KL_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/config/ResNet/ResNet50_vd-KL_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt new file mode 100644 index 0000000000000000000000000000000000000000..892b4859b1e3941c7b8ca235152f105d17d06729 --- /dev/null +++ b/test_tipc/config/ResNet/ResNet50_vd-KL_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt @@ -0,0 +1,18 @@ +===========================cpp_infer_params=========================== +model_name:ResNet50_vd_KL +cpp_infer_type:cls +cls_inference_model_dir:./ResNet50_vd_kl_quant_infer/ +det_inference_model_dir: +cls_inference_url:https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/inference/slim_model/ResNet50_vd_kl_quant_infer.tar +det_inference_url: +infer_quant:False +inference_cmd:./deploy/cpp/build/clas_system -c inference_cls.yaml +use_gpu:True|False +enable_mkldnn:False +cpu_threads:1 +batch_size:1 +use_tensorrt:False +precision:fp32 +image_dir:./dataset/ILSVRC2012/val/ILSVRC2012_val_00000001.JPEG +benchmark:False +generate_yaml_cmd:python3.7 test_tipc/generate_cpp_yaml.py diff --git a/test_tipc/config/ResNet/ResNet50_vd-KL_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/config/ResNet/ResNet50_vd-KL_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt new file mode 100644 index 0000000000000000000000000000000000000000..3e4b19b752e3f0f15fdb552a29035e055633102d --- /dev/null +++ b/test_tipc/config/ResNet/ResNet50_vd-KL_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt @@ -0,0 +1,14 @@ +===========================serving_params=========================== +model_name:ResNet50_vd_KL +python:python3.7 +inference_model_url:https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/inference/slim_model/ResNet50_vd_kl_quant_infer.tar +trans_model:-m paddle_serving_client.convert +--dirname:./deploy/paddleserving/ResNet50_vd_kl_quant_infer/ +--model_filename:inference.pdmodel +--params_filename:inference.pdiparams +--serving_server:./deploy/paddleserving/ResNet50_vd_kl_quant_serving/ +--serving_client:./deploy/paddleserving/ResNet50_vd_kl_quant_client/ +serving_dir:./deploy/paddleserving +web_service:null +--use_gpu:0|null +pipline:test_cpp_serving_client.py diff --git a/test_tipc/config/ResNet/ResNet50_vd-KL_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/config/ResNet/ResNet50_vd-KL_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt new file mode 100644 index 0000000000000000000000000000000000000000..236b57b03643d1e5cab743abb80dcf8cf03dc8c1 --- /dev/null +++ b/test_tipc/config/ResNet/ResNet50_vd-KL_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt @@ -0,0 +1,14 @@ +===========================serving_params=========================== +model_name:ResNet50_vd_KL +python:python3.7 +inference_model_url:https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/inference/slim_model/ResNet50_vd_kl_quant_infer.tar +trans_model:-m paddle_serving_client.convert +--dirname:./deploy/paddleserving/ResNet50_vd_kl_quant_infer/ +--model_filename:inference.pdmodel +--params_filename:inference.pdiparams +--serving_server:./deploy/paddleserving/ResNet50_vd_kl_quant_serving/ +--serving_client:./deploy/paddleserving/ResNet50_vd_kl_quant_client/ +serving_dir:./deploy/paddleserving +web_service:classification_web_service.py +--use_gpu:0|null +pipline:pipeline_http_client.py diff --git a/test_tipc/docs/test_inference_cpp.md b/test_tipc/docs/test_inference_cpp.md index 213959de0669f34dbb8bd51ad8fb78bff6e8e26c..e82b8ed800b4819e5f7599d565a0076137b8e531 100644 --- a/test_tipc/docs/test_inference_cpp.md +++ b/test_tipc/docs/test_inference_cpp.md @@ -6,25 +6,27 @@ Linux GPU/CPU C++ 推理功能测试的主程序为`test_inference_cpp.sh`，可 - 推理相关： -| 算法名称 | 模型名称 | device_CPU | device_GPU | -| :----: | :----: | :----: | :----: | -| MobileNetV3 | MobileNetV3_large_x1_0 | 支持 | 支持 | -| PP-ShiTu | PPShiTu_general_rec、PPShiTu_mainbody_det | 支持 | 支持 | -| PP-ShiTu | PPShiTu_mainbody_det | 支持 | 支持 | -| PPHGNet | PPHGNet_small | 支持 | 支持 | -| PPHGNet | PPHGNet_tiny | 支持 | 支持 | -| PPLCNet | PPLCNet_x0_25 | 支持 | 支持 | -| PPLCNet | PPLCNet_x0_35 | 支持 | 支持 | -| PPLCNet | PPLCNet_x0_5 | 支持 | 支持 | -| PPLCNet | PPLCNet_x0_75 | 支持 | 支持 | -| PPLCNet | PPLCNet_x1_0 | 支持 | 支持 | -| PPLCNet | PPLCNet_x1_5 | 支持 | 支持 | -| PPLCNet | PPLCNet_x2_0 | 支持 | 支持 | -| PPLCNet | PPLCNet_x2_5 | 支持 | 支持 | -| PPLCNetV2 | PPLCNetV2_base | 支持 | 支持 | -| ResNet | ResNet50 | 支持 | 支持 | -| ResNet | ResNet50_vd | 支持 | 支持 | -| SwinTransformer | SwinTransformer_tiny_patch4_window7_224 | 支持 | 支持 | +| 算法名称 | 模型名称 | device_CPU | device_GPU | +| :-------------: | :---------------------------------------: | :--------: | :--------: | +| MobileNetV3 | MobileNetV3_large_x1_0 | 支持 | 支持 | +| MobileNetV3 | MobileNetV3_large_x1_0_KL | 支持 | 支持 | +| PP-ShiTu | PPShiTu_general_rec、PPShiTu_mainbody_det | 支持 | 支持 | +| PP-ShiTu | PPShiTu_mainbody_det | 支持 | 支持 | +| PPHGNet | PPHGNet_small | 支持 | 支持 | +| PPHGNet | PPHGNet_tiny | 支持 | 支持 | +| PPLCNet | PPLCNet_x0_25 | 支持 | 支持 | +| PPLCNet | PPLCNet_x0_35 | 支持 | 支持 | +| PPLCNet | PPLCNet_x0_5 | 支持 | 支持 | +| PPLCNet | PPLCNet_x0_75 | 支持 | 支持 | +| PPLCNet | PPLCNet_x1_0 | 支持 | 支持 | +| PPLCNet | PPLCNet_x1_5 | 支持 | 支持 | +| PPLCNet | PPLCNet_x2_0 | 支持 | 支持 | +| PPLCNet | PPLCNet_x2_5 | 支持 | 支持 | +| PPLCNetV2 | PPLCNetV2_base | 支持 | 支持 | +| ResNet | ResNet50 | 支持 | 支持 | +| ResNet | ResNet50_vd | 支持 | 支持 | +| ResNet | ResNet50_vd_KL | 支持 | 支持 | +| SwinTransformer | SwinTransformer_tiny_patch4_window7_224 | 支持 | 支持 | ## 2. 测试流程(以**ResNet50**为例) @@ -167,11 +169,11 @@ build/paddle_inference_install_dir/ * [Paddle预测库官网](https://paddleinference.paddlepaddle.org.cn/user_guides/download_lib.html)上提供了不同cuda版本的Linux预测库，可以在官网查看并选择合适的预测库版本。 - 以`manylinux_cuda11.1_cudnn8.1_avx_mkl_trt7_gcc8.2`版本为例，使用下述命令下载并解压： + 以`manylinux_cuda10.1_cudnn7.6_avx_mkl_trt6_gcc8.2`版本为例，使用下述命令下载并解压： ```shell -wget https://paddle-inference-lib.bj.bcebos.com/2.2.2/cxx_c/Linux/GPU/x86-64_gcc8.2_avx_mkl_cuda11.1_cudnn8.1.1_trt7.2.3.4/paddle_inference.tgz +wget https://paddle-inference-lib.bj.bcebos.com/2.2.2/cxx_c/Linux/GPU/x86-64_gcc8.2_avx_mkl_cuda10.1_cudnn7.6.5_trt6.0.1.5/paddle_inference.tgz tar -xvf paddle_inference.tgz ``` diff --git a/test_tipc/docs/test_serving_infer_cpp.md b/test_tipc/docs/test_serving_infer_cpp.md index df13c2b26272dc1fdd4262f09f2e7bcd4978a114..2018f4c9215d91d2e4b006743e54293315fdf346 100644 --- a/test_tipc/docs/test_serving_infer_cpp.md +++ b/test_tipc/docs/test_serving_infer_cpp.md @@ -10,6 +10,7 @@ Linux GPU/CPU C++ 服务化部署测试的主程序为`test_serving_infer_cpp.sh | 算法名称 | 模型名称 | device_CPU | device_GPU | | :-------------: | :---------------------------------------: | :--------: | :--------: | | MobileNetV3 | MobileNetV3_large_x1_0 | 支持 | 支持 | +| MobileNetV3 | MobileNetV3_large_x1_0_KL | 支持 | 支持 | | PP-ShiTu | PPShiTu_general_rec、PPShiTu_mainbody_det | 支持 | 支持 | | PPHGNet | PPHGNet_small | 支持 | 支持 | | PPHGNet | PPHGNet_tiny | 支持 | 支持 | @@ -24,6 +25,7 @@ Linux GPU/CPU C++ 服务化部署测试的主程序为`test_serving_infer_cpp.sh | PPLCNetV2 | PPLCNetV2_base | 支持 | 支持 | | ResNet | ResNet50 | 支持 | 支持 | | ResNet | ResNet50_vd | 支持 | 支持 | +| ResNet | ResNet50_vd_KL | 支持 | 支持 | | SwinTransformer | SwinTransformer_tiny_patch4_window7_224 | 支持 | 支持 | diff --git a/test_tipc/docs/test_serving_infer_python.md b/test_tipc/docs/test_serving_infer_python.md index 1e802d3ff4fa11b70dedd603a8deea8fd7ab846d..1ba4c1516dc07f2a88aa406c9316e871fd7c8a8b 100644 --- a/test_tipc/docs/test_serving_infer_python.md +++ b/test_tipc/docs/test_serving_infer_python.md @@ -10,6 +10,7 @@ Linux GPU/CPU PYTHON 服务化部署测试的主程序为`test_serving_infer_pyt | 算法名称 | 模型名称 | device_CPU | device_GPU | | :-------------: | :---------------------------------------: | :--------: | :--------: | | MobileNetV3 | MobileNetV3_large_x1_0 | 支持 | 支持 | +| MobileNetV3 | MobileNetV3_large_x1_0_KL | 支持 | 支持 | | PP-ShiTu | PPShiTu_general_rec、PPShiTu_mainbody_det | 支持 | 支持 | | PPHGNet | PPHGNet_small | 支持 | 支持 | | PPHGNet | PPHGNet_tiny | 支持 | 支持 | @@ -24,6 +25,7 @@ Linux GPU/CPU PYTHON 服务化部署测试的主程序为`test_serving_infer_pyt | PPLCNetV2 | PPLCNetV2_base | 支持 | 支持 | | ResNet | ResNet50 | 支持 | 支持 | | ResNet | ResNet50_vd | 支持 | 支持 | +| ResNet | ResNet50_vd_KL | 支持 | 支持 | | SwinTransformer | SwinTransformer_tiny_patch4_window7_224 | 支持 | 支持 | diff --git a/test_tipc/prepare.sh b/test_tipc/prepare.sh index e0908d76c5f41fd32d8c1d6a282e7fe01d675753..f2e48710d884a07d0282cfaac225b1b9f7a31dcd 100644 --- a/test_tipc/prepare.sh +++ b/test_tipc/prepare.sh @@ -83,6 +83,12 @@ if [[ ${MODE} = "cpp_infer" ]]; then popd echo "################### build opencv finished ###################" fi + if [[ ! -d "./deploy/cpp/paddle_inference/" ]]; then + pushd ./deploy/cpp/ + wget -nc https://paddle-inference-lib.bj.bcebos.com/2.2.2/cxx_c/Linux/GPU/x86-64_gcc8.2_avx_mkl_cuda10.1_cudnn7.6.5_trt6.0.1.5/paddle_inference.tgz + tar xf paddle_inference.tgz + popd + fi if [[ $FILENAME == *infer_cpp_linux_gpu_cpu.txt ]]; then cpp_type=$(func_parser_value "${lines[2]}") cls_inference_model_dir=$(func_parser_value "${lines[3]}")