add kl_quant chain and polish prepare.sh

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)
 
 

test_tipc/config/MobileNetV3/MobileNetV3_large_x1_0-KL_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt

+===========================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

test_tipc/config/ResNet/ResNet50_vd-KL_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt

+===========================cpp_infer_params===========================
+model_name:ResNet50_vd_kl_quant
+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

test_tipc/docs/test_inference_cpp.md

@@ -167,11 +167,11 @@ 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
 ```

test_tipc/prepare.sh

@@ -50,6 +50,8 @@ if [[ ${MODE} = "cpp_infer" ]]; then
         echo "################### build opencv ###################"
         rm -rf ./deploy/cpp/opencv-3.4.7.tar.gz ./deploy/cpp/opencv-3.4.7/
         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
         wget -nc https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/opencv-3.4.7.tar.gz
         tar -xf opencv-3.4.7.tar.gz