未验证 提交 45f647db 编写于 作者: D Double_V 提交者: GitHub

Merge pull request #1542 from LDOUBLEV/cpp_2.0

update cpp_infer to 2.0
...@@ -25,9 +25,9 @@ ...@@ -25,9 +25,9 @@
namespace PaddleOCR { namespace PaddleOCR {
class Config { class OCRConfig {
public: public:
explicit Config(const std::string &config_file) { explicit OCRConfig(const std::string &config_file) {
config_map_ = LoadConfig(config_file); config_map_ = LoadConfig(config_file);
this->use_gpu = bool(stoi(config_map_["use_gpu"])); this->use_gpu = bool(stoi(config_map_["use_gpu"]));
...@@ -41,8 +41,6 @@ public: ...@@ -41,8 +41,6 @@ public:
this->use_mkldnn = bool(stoi(config_map_["use_mkldnn"])); this->use_mkldnn = bool(stoi(config_map_["use_mkldnn"]));
this->use_zero_copy_run = bool(stoi(config_map_["use_zero_copy_run"]));
this->max_side_len = stoi(config_map_["max_side_len"]); this->max_side_len = stoi(config_map_["max_side_len"]);
this->det_db_thresh = stod(config_map_["det_db_thresh"]); this->det_db_thresh = stod(config_map_["det_db_thresh"]);
...@@ -76,8 +74,6 @@ public: ...@@ -76,8 +74,6 @@ public:
bool use_mkldnn = false; bool use_mkldnn = false;
bool use_zero_copy_run = false;
int max_side_len = 960; int max_side_len = 960;
double det_db_thresh = 0.3; double det_db_thresh = 0.3;
......
...@@ -30,6 +30,8 @@ ...@@ -30,6 +30,8 @@
#include <include/preprocess_op.h> #include <include/preprocess_op.h>
#include <include/utility.h> #include <include/utility.h>
using namespace paddle_infer;
namespace PaddleOCR { namespace PaddleOCR {
class Classifier { class Classifier {
...@@ -37,14 +39,12 @@ public: ...@@ -37,14 +39,12 @@ public:
explicit Classifier(const std::string &model_dir, const bool &use_gpu, explicit Classifier(const std::string &model_dir, const bool &use_gpu,
const int &gpu_id, const int &gpu_mem, const int &gpu_id, const int &gpu_mem,
const int &cpu_math_library_num_threads, const int &cpu_math_library_num_threads,
const bool &use_mkldnn, const bool &use_zero_copy_run, const bool &use_mkldnn, const double &cls_thresh) {
const double &cls_thresh) {
this->use_gpu_ = use_gpu; this->use_gpu_ = use_gpu;
this->gpu_id_ = gpu_id; this->gpu_id_ = gpu_id;
this->gpu_mem_ = gpu_mem; this->gpu_mem_ = gpu_mem;
this->cpu_math_library_num_threads_ = cpu_math_library_num_threads; this->cpu_math_library_num_threads_ = cpu_math_library_num_threads;
this->use_mkldnn_ = use_mkldnn; this->use_mkldnn_ = use_mkldnn;
this->use_zero_copy_run_ = use_zero_copy_run;
this->cls_thresh = cls_thresh; this->cls_thresh = cls_thresh;
...@@ -57,14 +57,13 @@ public: ...@@ -57,14 +57,13 @@ public:
cv::Mat Run(cv::Mat &img); cv::Mat Run(cv::Mat &img);
private: private:
std::shared_ptr<PaddlePredictor> predictor_; std::shared_ptr<Predictor> predictor_;
bool use_gpu_ = false; bool use_gpu_ = false;
int gpu_id_ = 0; int gpu_id_ = 0;
int gpu_mem_ = 4000; int gpu_mem_ = 4000;
int cpu_math_library_num_threads_ = 4; int cpu_math_library_num_threads_ = 4;
bool use_mkldnn_ = false; bool use_mkldnn_ = false;
bool use_zero_copy_run_ = false;
double cls_thresh = 0.5; double cls_thresh = 0.5;
std::vector<float> mean_ = {0.5f, 0.5f, 0.5f}; std::vector<float> mean_ = {0.5f, 0.5f, 0.5f};
......
...@@ -32,6 +32,8 @@ ...@@ -32,6 +32,8 @@
#include <include/postprocess_op.h> #include <include/postprocess_op.h>
#include <include/preprocess_op.h> #include <include/preprocess_op.h>
using namespace paddle_infer;
namespace PaddleOCR { namespace PaddleOCR {
class DBDetector { class DBDetector {
...@@ -39,8 +41,8 @@ public: ...@@ -39,8 +41,8 @@ public:
explicit DBDetector(const std::string &model_dir, const bool &use_gpu, explicit DBDetector(const std::string &model_dir, const bool &use_gpu,
const int &gpu_id, const int &gpu_mem, const int &gpu_id, const int &gpu_mem,
const int &cpu_math_library_num_threads, const int &cpu_math_library_num_threads,
const bool &use_mkldnn, const bool &use_zero_copy_run, const bool &use_mkldnn, const int &max_side_len,
const int &max_side_len, const double &det_db_thresh, const double &det_db_thresh,
const double &det_db_box_thresh, const double &det_db_box_thresh,
const double &det_db_unclip_ratio, const double &det_db_unclip_ratio,
const bool &visualize) { const bool &visualize) {
...@@ -49,7 +51,6 @@ public: ...@@ -49,7 +51,6 @@ public:
this->gpu_mem_ = gpu_mem; this->gpu_mem_ = gpu_mem;
this->cpu_math_library_num_threads_ = cpu_math_library_num_threads; this->cpu_math_library_num_threads_ = cpu_math_library_num_threads;
this->use_mkldnn_ = use_mkldnn; this->use_mkldnn_ = use_mkldnn;
this->use_zero_copy_run_ = use_zero_copy_run;
this->max_side_len_ = max_side_len; this->max_side_len_ = max_side_len;
...@@ -69,14 +70,13 @@ public: ...@@ -69,14 +70,13 @@ public:
void Run(cv::Mat &img, std::vector<std::vector<std::vector<int>>> &boxes); void Run(cv::Mat &img, std::vector<std::vector<std::vector<int>>> &boxes);
private: private:
std::shared_ptr<PaddlePredictor> predictor_; std::shared_ptr<Predictor> predictor_;
bool use_gpu_ = false; bool use_gpu_ = false;
int gpu_id_ = 0; int gpu_id_ = 0;
int gpu_mem_ = 4000; int gpu_mem_ = 4000;
int cpu_math_library_num_threads_ = 4; int cpu_math_library_num_threads_ = 4;
bool use_mkldnn_ = false; bool use_mkldnn_ = false;
bool use_zero_copy_run_ = false;
int max_side_len_ = 960; int max_side_len_ = 960;
......
...@@ -32,6 +32,8 @@ ...@@ -32,6 +32,8 @@
#include <include/preprocess_op.h> #include <include/preprocess_op.h>
#include <include/utility.h> #include <include/utility.h>
using namespace paddle_infer;
namespace PaddleOCR { namespace PaddleOCR {
class CRNNRecognizer { class CRNNRecognizer {
...@@ -39,14 +41,12 @@ public: ...@@ -39,14 +41,12 @@ public:
explicit CRNNRecognizer(const std::string &model_dir, const bool &use_gpu, explicit CRNNRecognizer(const std::string &model_dir, const bool &use_gpu,
const int &gpu_id, const int &gpu_mem, const int &gpu_id, const int &gpu_mem,
const int &cpu_math_library_num_threads, const int &cpu_math_library_num_threads,
const bool &use_mkldnn, const bool &use_zero_copy_run, const bool &use_mkldnn, const string &label_path) {
const string &label_path) {
this->use_gpu_ = use_gpu; this->use_gpu_ = use_gpu;
this->gpu_id_ = gpu_id; this->gpu_id_ = gpu_id;
this->gpu_mem_ = gpu_mem; this->gpu_mem_ = gpu_mem;
this->cpu_math_library_num_threads_ = cpu_math_library_num_threads; this->cpu_math_library_num_threads_ = cpu_math_library_num_threads;
this->use_mkldnn_ = use_mkldnn; this->use_mkldnn_ = use_mkldnn;
this->use_zero_copy_run_ = use_zero_copy_run;
this->label_list_ = Utility::ReadDict(label_path); this->label_list_ = Utility::ReadDict(label_path);
this->label_list_.insert(this->label_list_.begin(), this->label_list_.insert(this->label_list_.begin(),
...@@ -63,14 +63,13 @@ public: ...@@ -63,14 +63,13 @@ public:
Classifier *cls); Classifier *cls);
private: private:
std::shared_ptr<PaddlePredictor> predictor_; std::shared_ptr<Predictor> predictor_;
bool use_gpu_ = false; bool use_gpu_ = false;
int gpu_id_ = 0; int gpu_id_ = 0;
int gpu_mem_ = 4000; int gpu_mem_ = 4000;
int cpu_math_library_num_threads_ = 4; int cpu_math_library_num_threads_ = 4;
bool use_mkldnn_ = false; bool use_mkldnn_ = false;
bool use_zero_copy_run_ = false;
std::vector<std::string> label_list_; std::vector<std::string> label_list_;
......
...@@ -16,7 +16,7 @@ ...@@ -16,7 +16,7 @@
namespace PaddleOCR { namespace PaddleOCR {
std::vector<std::string> Config::split(const std::string &str, std::vector<std::string> OCRConfig::split(const std::string &str,
const std::string &delim) { const std::string &delim) {
std::vector<std::string> res; std::vector<std::string> res;
if ("" == str) if ("" == str)
...@@ -38,7 +38,7 @@ std::vector<std::string> Config::split(const std::string &str, ...@@ -38,7 +38,7 @@ std::vector<std::string> Config::split(const std::string &str,
} }
std::map<std::string, std::string> std::map<std::string, std::string>
Config::LoadConfig(const std::string &config_path) { OCRConfig::LoadConfig(const std::string &config_path) {
auto config = Utility::ReadDict(config_path); auto config = Utility::ReadDict(config_path);
std::map<std::string, std::string> dict; std::map<std::string, std::string> dict;
...@@ -53,7 +53,7 @@ Config::LoadConfig(const std::string &config_path) { ...@@ -53,7 +53,7 @@ Config::LoadConfig(const std::string &config_path) {
return dict; return dict;
} }
void Config::PrintConfigInfo() { void OCRConfig::PrintConfigInfo() {
std::cout << "=======Paddle OCR inference config======" << std::endl; std::cout << "=======Paddle OCR inference config======" << std::endl;
for (auto iter = config_map_.begin(); iter != config_map_.end(); iter++) { for (auto iter = config_map_.begin(); iter != config_map_.end(); iter++) {
std::cout << iter->first << " : " << iter->second << std::endl; std::cout << iter->first << " : " << iter->second << std::endl;
......
...@@ -42,7 +42,7 @@ int main(int argc, char **argv) { ...@@ -42,7 +42,7 @@ int main(int argc, char **argv) {
exit(1); exit(1);
} }
Config config(argv[1]); OCRConfig config(argv[1]);
config.PrintConfigInfo(); config.PrintConfigInfo();
...@@ -50,37 +50,22 @@ int main(int argc, char **argv) { ...@@ -50,37 +50,22 @@ int main(int argc, char **argv) {
cv::Mat srcimg = cv::imread(img_path, cv::IMREAD_COLOR); cv::Mat srcimg = cv::imread(img_path, cv::IMREAD_COLOR);
DBDetector det( DBDetector det(config.det_model_dir, config.use_gpu, config.gpu_id,
config.det_model_dir, config.use_gpu, config.gpu_id, config.gpu_mem, config.gpu_mem, config.cpu_math_library_num_threads,
config.cpu_math_library_num_threads, config.use_mkldnn, config.use_mkldnn, config.max_side_len, config.det_db_thresh,
config.use_zero_copy_run, config.max_side_len, config.det_db_thresh, config.det_db_box_thresh, config.det_db_unclip_ratio,
config.det_db_box_thresh, config.det_db_unclip_ratio, config.visualize); config.visualize);
Classifier *cls = nullptr; Classifier *cls = nullptr;
if (config.use_angle_cls == true) { if (config.use_angle_cls == true) {
cls = new Classifier(config.cls_model_dir, config.use_gpu, config.gpu_id, cls = new Classifier(config.cls_model_dir, config.use_gpu, config.gpu_id,
config.gpu_mem, config.cpu_math_library_num_threads, config.gpu_mem, config.cpu_math_library_num_threads,
config.use_mkldnn, config.use_zero_copy_run, config.use_mkldnn, config.cls_thresh);
config.cls_thresh);
} }
CRNNRecognizer rec(config.rec_model_dir, config.use_gpu, config.gpu_id, CRNNRecognizer rec(config.rec_model_dir, config.use_gpu, config.gpu_id,
config.gpu_mem, config.cpu_math_library_num_threads, config.gpu_mem, config.cpu_math_library_num_threads,
config.use_mkldnn, config.use_zero_copy_run, config.use_mkldnn, config.char_list_file);
config.char_list_file);
#ifdef USE_MKL
#pragma omp parallel
for (auto i = 0; i < 10; i++) {
LOG_IF(WARNING,
config.cpu_math_library_num_threads != omp_get_num_threads())
<< "WARNING! MKL is running on " << omp_get_num_threads()
<< " threads while cpu_math_library_num_threads is set to "
<< config.cpu_math_library_num_threads
<< ". Possible reason could be 1. You have set omp_set_num_threads() "
"somewhere; 2. MKL is not linked properly";
}
#endif
auto start = std::chrono::system_clock::now(); auto start = std::chrono::system_clock::now();
std::vector<std::vector<std::vector<int>>> boxes; std::vector<std::vector<std::vector<int>>> boxes;
......
...@@ -35,26 +35,16 @@ cv::Mat Classifier::Run(cv::Mat &img) { ...@@ -35,26 +35,16 @@ cv::Mat Classifier::Run(cv::Mat &img) {
this->permute_op_.Run(&resize_img, input.data()); this->permute_op_.Run(&resize_img, input.data());
// Inference. // Inference.
if (this->use_zero_copy_run_) {
auto input_names = this->predictor_->GetInputNames(); auto input_names = this->predictor_->GetInputNames();
auto input_t = this->predictor_->GetInputTensor(input_names[0]); auto input_t = this->predictor_->GetInputHandle(input_names[0]);
input_t->Reshape({1, 3, resize_img.rows, resize_img.cols}); input_t->Reshape({1, 3, resize_img.rows, resize_img.cols});
input_t->copy_from_cpu(input.data()); input_t->CopyFromCpu(input.data());
this->predictor_->ZeroCopyRun(); this->predictor_->Run();
} else {
paddle::PaddleTensor input_t;
input_t.shape = {1, 3, resize_img.rows, resize_img.cols};
input_t.data =
paddle::PaddleBuf(input.data(), input.size() * sizeof(float));
input_t.dtype = PaddleDType::FLOAT32;
std::vector<paddle::PaddleTensor> outputs;
this->predictor_->Run({input_t}, &outputs, 1);
}
std::vector<float> softmax_out; std::vector<float> softmax_out;
std::vector<int64_t> label_out; std::vector<int64_t> label_out;
auto output_names = this->predictor_->GetOutputNames(); auto output_names = this->predictor_->GetOutputNames();
auto softmax_out_t = this->predictor_->GetOutputTensor(output_names[0]); auto softmax_out_t = this->predictor_->GetOutputHandle(output_names[0]);
auto softmax_shape_out = softmax_out_t->shape(); auto softmax_shape_out = softmax_out_t->shape();
int softmax_out_num = int softmax_out_num =
...@@ -63,7 +53,7 @@ cv::Mat Classifier::Run(cv::Mat &img) { ...@@ -63,7 +53,7 @@ cv::Mat Classifier::Run(cv::Mat &img) {
softmax_out.resize(softmax_out_num); softmax_out.resize(softmax_out_num);
softmax_out_t->copy_to_cpu(softmax_out.data()); softmax_out_t->CopyToCpu(softmax_out.data());
float score = 0; float score = 0;
int label = 0; int label = 0;
...@@ -95,7 +85,7 @@ void Classifier::LoadModel(const std::string &model_dir) { ...@@ -95,7 +85,7 @@ void Classifier::LoadModel(const std::string &model_dir) {
} }
// false for zero copy tensor // false for zero copy tensor
config.SwitchUseFeedFetchOps(!this->use_zero_copy_run_); config.SwitchUseFeedFetchOps(false);
// true for multiple input // true for multiple input
config.SwitchSpecifyInputNames(true); config.SwitchSpecifyInputNames(true);
...@@ -104,6 +94,6 @@ void Classifier::LoadModel(const std::string &model_dir) { ...@@ -104,6 +94,6 @@ void Classifier::LoadModel(const std::string &model_dir) {
config.EnableMemoryOptim(); config.EnableMemoryOptim();
config.DisableGlogInfo(); config.DisableGlogInfo();
this->predictor_ = CreatePaddlePredictor(config); this->predictor_ = CreatePredictor(config);
} }
} // namespace PaddleOCR } // namespace PaddleOCR
...@@ -17,12 +17,17 @@ ...@@ -17,12 +17,17 @@
namespace PaddleOCR { namespace PaddleOCR {
void DBDetector::LoadModel(const std::string &model_dir) { void DBDetector::LoadModel(const std::string &model_dir) {
AnalysisConfig config; // AnalysisConfig config;
paddle_infer::Config config;
config.SetModel(model_dir + "/inference.pdmodel", config.SetModel(model_dir + "/inference.pdmodel",
model_dir + "/inference.pdiparams"); model_dir + "/inference.pdiparams");
if (this->use_gpu_) { if (this->use_gpu_) {
config.EnableUseGpu(this->gpu_mem_, this->gpu_id_); config.EnableUseGpu(this->gpu_mem_, this->gpu_id_);
// config.EnableTensorRtEngine(
// 1 << 20, 1, 3,
// AnalysisConfig::Precision::kFloat32,
// false, false);
} else { } else {
config.DisableGpu(); config.DisableGpu();
if (this->use_mkldnn_) { if (this->use_mkldnn_) {
...@@ -32,10 +37,8 @@ void DBDetector::LoadModel(const std::string &model_dir) { ...@@ -32,10 +37,8 @@ void DBDetector::LoadModel(const std::string &model_dir) {
} }
config.SetCpuMathLibraryNumThreads(this->cpu_math_library_num_threads_); config.SetCpuMathLibraryNumThreads(this->cpu_math_library_num_threads_);
} }
// use zero_copy_run as default
// false for zero copy tensor config.SwitchUseFeedFetchOps(false);
// true for commom tensor
config.SwitchUseFeedFetchOps(!this->use_zero_copy_run_);
// true for multiple input // true for multiple input
config.SwitchSpecifyInputNames(true); config.SwitchSpecifyInputNames(true);
...@@ -44,7 +47,7 @@ void DBDetector::LoadModel(const std::string &model_dir) { ...@@ -44,7 +47,7 @@ void DBDetector::LoadModel(const std::string &model_dir) {
config.EnableMemoryOptim(); config.EnableMemoryOptim();
config.DisableGlogInfo(); config.DisableGlogInfo();
this->predictor_ = CreatePaddlePredictor(config); this->predictor_ = CreatePredictor(config);
} }
void DBDetector::Run(cv::Mat &img, void DBDetector::Run(cv::Mat &img,
...@@ -64,31 +67,21 @@ void DBDetector::Run(cv::Mat &img, ...@@ -64,31 +67,21 @@ void DBDetector::Run(cv::Mat &img,
this->permute_op_.Run(&resize_img, input.data()); this->permute_op_.Run(&resize_img, input.data());
// Inference. // Inference.
if (this->use_zero_copy_run_) {
auto input_names = this->predictor_->GetInputNames(); auto input_names = this->predictor_->GetInputNames();
auto input_t = this->predictor_->GetInputTensor(input_names[0]); auto input_t = this->predictor_->GetInputHandle(input_names[0]);
input_t->Reshape({1, 3, resize_img.rows, resize_img.cols}); input_t->Reshape({1, 3, resize_img.rows, resize_img.cols});
input_t->copy_from_cpu(input.data()); input_t->CopyFromCpu(input.data());
this->predictor_->ZeroCopyRun(); this->predictor_->Run();
} else {
paddle::PaddleTensor input_t;
input_t.shape = {1, 3, resize_img.rows, resize_img.cols};
input_t.data =
paddle::PaddleBuf(input.data(), input.size() * sizeof(float));
input_t.dtype = PaddleDType::FLOAT32;
std::vector<paddle::PaddleTensor> outputs;
this->predictor_->Run({input_t}, &outputs, 1);
}
std::vector<float> out_data; std::vector<float> out_data;
auto output_names = this->predictor_->GetOutputNames(); auto output_names = this->predictor_->GetOutputNames();
auto output_t = this->predictor_->GetOutputTensor(output_names[0]); auto output_t = this->predictor_->GetOutputHandle(output_names[0]);
std::vector<int> output_shape = output_t->shape(); std::vector<int> output_shape = output_t->shape();
int out_num = std::accumulate(output_shape.begin(), output_shape.end(), 1, int out_num = std::accumulate(output_shape.begin(), output_shape.end(), 1,
std::multiplies<int>()); std::multiplies<int>());
out_data.resize(out_num); out_data.resize(out_num);
output_t->copy_to_cpu(out_data.data()); output_t->CopyToCpu(out_data.data());
int n2 = output_shape[2]; int n2 = output_shape[2];
int n3 = output_shape[3]; int n3 = output_shape[3];
......
...@@ -43,32 +43,22 @@ void CRNNRecognizer::Run(std::vector<std::vector<std::vector<int>>> boxes, ...@@ -43,32 +43,22 @@ void CRNNRecognizer::Run(std::vector<std::vector<std::vector<int>>> boxes,
this->permute_op_.Run(&resize_img, input.data()); this->permute_op_.Run(&resize_img, input.data());
// Inference. // Inference.
if (this->use_zero_copy_run_) {
auto input_names = this->predictor_->GetInputNames(); auto input_names = this->predictor_->GetInputNames();
auto input_t = this->predictor_->GetInputTensor(input_names[0]); auto input_t = this->predictor_->GetInputHandle(input_names[0]);
input_t->Reshape({1, 3, resize_img.rows, resize_img.cols}); input_t->Reshape({1, 3, resize_img.rows, resize_img.cols});
input_t->copy_from_cpu(input.data()); input_t->CopyFromCpu(input.data());
this->predictor_->ZeroCopyRun(); this->predictor_->Run();
} else {
paddle::PaddleTensor input_t;
input_t.shape = {1, 3, resize_img.rows, resize_img.cols};
input_t.data =
paddle::PaddleBuf(input.data(), input.size() * sizeof(float));
input_t.dtype = PaddleDType::FLOAT32;
std::vector<paddle::PaddleTensor> outputs;
this->predictor_->Run({input_t}, &outputs, 1);
}
std::vector<float> predict_batch; std::vector<float> predict_batch;
auto output_names = this->predictor_->GetOutputNames(); auto output_names = this->predictor_->GetOutputNames();
auto output_t = this->predictor_->GetOutputTensor(output_names[0]); auto output_t = this->predictor_->GetOutputHandle(output_names[0]);
auto predict_shape = output_t->shape(); auto predict_shape = output_t->shape();
int out_num = std::accumulate(predict_shape.begin(), predict_shape.end(), 1, int out_num = std::accumulate(predict_shape.begin(), predict_shape.end(), 1,
std::multiplies<int>()); std::multiplies<int>());
predict_batch.resize(out_num); predict_batch.resize(out_num);
output_t->copy_to_cpu(predict_batch.data()); output_t->CopyToCpu(predict_batch.data());
// ctc decode // ctc decode
std::vector<std::string> str_res; std::vector<std::string> str_res;
...@@ -102,7 +92,8 @@ void CRNNRecognizer::Run(std::vector<std::vector<std::vector<int>>> boxes, ...@@ -102,7 +92,8 @@ void CRNNRecognizer::Run(std::vector<std::vector<std::vector<int>>> boxes,
} }
void CRNNRecognizer::LoadModel(const std::string &model_dir) { void CRNNRecognizer::LoadModel(const std::string &model_dir) {
AnalysisConfig config; // AnalysisConfig config;
paddle_infer::Config config;
config.SetModel(model_dir + "/inference.pdmodel", config.SetModel(model_dir + "/inference.pdmodel",
model_dir + "/inference.pdiparams"); model_dir + "/inference.pdiparams");
...@@ -118,9 +109,7 @@ void CRNNRecognizer::LoadModel(const std::string &model_dir) { ...@@ -118,9 +109,7 @@ void CRNNRecognizer::LoadModel(const std::string &model_dir) {
config.SetCpuMathLibraryNumThreads(this->cpu_math_library_num_threads_); config.SetCpuMathLibraryNumThreads(this->cpu_math_library_num_threads_);
} }
// false for zero copy tensor config.SwitchUseFeedFetchOps(false);
// true for commom tensor
config.SwitchUseFeedFetchOps(!this->use_zero_copy_run_);
// true for multiple input // true for multiple input
config.SwitchSpecifyInputNames(true); config.SwitchSpecifyInputNames(true);
...@@ -129,7 +118,7 @@ void CRNNRecognizer::LoadModel(const std::string &model_dir) { ...@@ -129,7 +118,7 @@ void CRNNRecognizer::LoadModel(const std::string &model_dir) {
config.EnableMemoryOptim(); config.EnableMemoryOptim();
config.DisableGlogInfo(); config.DisableGlogInfo();
this->predictor_ = CreatePaddlePredictor(config); this->predictor_ = CreatePredictor(config);
} }
cv::Mat CRNNRecognizer::GetRotateCropImage(const cv::Mat &srcimage, cv::Mat CRNNRecognizer::GetRotateCropImage(const cv::Mat &srcimage,
......
...@@ -4,7 +4,6 @@ gpu_id 0 ...@@ -4,7 +4,6 @@ gpu_id 0
gpu_mem 4000 gpu_mem 4000
cpu_math_library_num_threads 10 cpu_math_library_num_threads 10
use_mkldnn 0 use_mkldnn 0
use_zero_copy_run 1
# det config # det config
max_side_len 960 max_side_len 960
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册