Skip to content

S
Serving

PaddlePaddle / Serving
大约 1 年 前同步成功

通知 185
Star 833
Fork 253
S
Serving

体验新版 GitCode，发现更多精彩内容 >>

提交 e832b55d 编写于 作者: W wangjiawei04
浏览文件
操作

rpc local unifed

上级 e54b5ad9 21c19e65
隐藏空白更改
内联 并排
Showing with 361 addition and 216 deletion
core/configure/proto/server_configure.proto
浏览文件 @ e832b55d
......@@ -58,6 +58,8 @@ message ResourceConf {
optional string cube_config_path = 5;
optional string cube_config_file = 6;
optional int32 cube_quant_bits = 7; // set 0 if no quant.
optional string auth_product_name = 8;
optional string auth_container_id = 9;
};
// DAG node depency info
......
core/cube/cube-api/include/meta.h
浏览文件 @ e832b55d
......@@ -22,7 +22,8 @@
#ifdef BCLOUD
#include "baidu/rpc/channel.h"
#include "baidu/rpc/parallel_channel.h"
#include "rapidjson/document.h"
#include "rapidjson_1.0/document.h"
#include "rapidjson_1.0/rapidjson.h"
#else
#include "brpc/channel.h"
#include "brpc/parallel_channel.h"
......
core/general-client/include/general_model.h
浏览文件 @ e832b55d
......@@ -217,7 +217,7 @@ class PredictorClient {
int create_predictor();
int destroy_predictor();
int numpy_predict(
const std::vector<std::vector<py::array_t<float>>>& float_feed_batch,
const std::vector<std::string>& float_feed_name,
......
core/general-client/src/general_model.cpp
浏览文件 @ e832b55d
......@@ -39,7 +39,9 @@ using configure::GeneralModelConfig;
void PredictorClient::init_gflags(std::vector<std::string> argv) {
std::call_once(gflags_init_flag, [&]() {
#ifndef BCLOUD
FLAGS_logtostderr = true;
#endif
argv.insert(argv.begin(), "dummy");
int argc = argv.size();
char **arr = new char *[argv.size()];
......@@ -135,7 +137,6 @@ int PredictorClient::create_predictor() {
return 0;
}
int PredictorClient::numpy_predict(
const std::vector<std::vector<py::array_t<float>>> &float_feed_batch,
const std::vector<std::string> &float_feed_name,
......@@ -143,7 +144,7 @@ int PredictorClient::numpy_predict(
const std::vector<std::vector<py::array_t<int64_t>>> &int_feed_batch,
const std::vector<std::string> &int_feed_name,
const std::vector<std::vector<int>> &int_shape,
const std::vector<std::vector<int>>& lod_slot_batch,
const std::vector<std::vector<int>> &lod_slot_batch,
const std::vector<std::string> &fetch_name,
PredictorRes &predict_res_batch,
const int &pid,
......@@ -197,7 +198,8 @@ int PredictorClient::numpy_predict(
<< float_shape[vec_idx].size();
for (uint32_t j = 0; j < float_shape[vec_idx].size(); ++j) {
tensor->add_shape(float_shape[vec_idx][j]);
std::cout << "shape " << j << " : " << float_shape[vec_idx][j] << std::endl;
std::cout << "shape " << j << " : " << float_shape[vec_idx][j]
<< std::endl;
}
for (uint32_t j = 0; j < lod_slot_batch[vec_idx].size(); ++j) {
tensor->add_lod(lod_slot_batch[vec_idx][j]);
......
core/general-client/src/pybind_general_model.cpp
浏览文件 @ e832b55d
......@@ -105,7 +105,7 @@ PYBIND11_MODULE(serving_client, m) {
&int_feed_batch,
const std::vector<std::string> &int_feed_name,
const std::vector<std::vector<int>> &int_shape,
const std::vector<std::vector<int>>& lod_slot_batch,
const std::vector<std::vector<int>> &lod_slot_batch,
const std::vector<std::string> &fetch_name,
PredictorRes &predict_res_batch,
const int &pid,
......
core/general-server/op/general_copy_op.h
浏览文件 @ e832b55d
......@@ -13,20 +13,12 @@
// limitations under the License.
#pragma once
#include <vector>
#ifdef BCLOUD
#ifdef WITH_GPU
#include "paddle/paddle_inference_api.h"
#else
#include "paddle/fluid/inference/api/paddle_inference_api.h"
#endif
#else
#include "paddle_inference_api.h" // NOLINT
#endif
#include <string>
#include <vector>
#include "core/general-server/general_model_service.pb.h"
#include "core/general-server/op/general_infer_helper.h"
#include "core/predictor/framework/resource.h"
#include "paddle_inference_api.h" // NOLINT
namespace baidu {
namespace paddle_serving {
......
core/general-server/op/general_dist_kv_infer_op.h
浏览文件 @ e832b55d
......@@ -15,17 +15,9 @@
#pragma once
#include <string>
#include <vector>
#ifdef BCLOUD
#ifdef WITH_GPU
#include "paddle/paddle_inference_api.h"
#else
#include "paddle/fluid/inference/api/paddle_inference_api.h"
#endif
#else
#include "paddle_inference_api.h" // NOLINT
#endif
#include "core/general-server/general_model_service.pb.h"
#include "core/general-server/op/general_infer_helper.h"
#include "paddle_inference_api.h" // NOLINT
namespace baidu {
namespace paddle_serving {
......
core/general-server/op/general_dist_kv_quant_infer_op.h
浏览文件 @ e832b55d
......@@ -15,17 +15,9 @@
#pragma once
#include <string>
#include <vector>
#ifdef BCLOUD
#ifdef WITH_GPU
#include "paddle/paddle_inference_api.h"
#else
#include "paddle/fluid/inference/api/paddle_inference_api.h"
#endif
#else
#include "paddle_inference_api.h" // NOLINT
#endif
#include "core/general-server/general_model_service.pb.h"
#include "core/general-server/op/general_infer_helper.h"
#include "paddle_inference_api.h" // NOLINT
namespace baidu {
namespace paddle_serving {
......
core/general-server/op/general_infer_helper.h
浏览文件 @ e832b55d
......@@ -15,17 +15,9 @@
#pragma once
#include <string.h>
#include <string>
#include <vector>
#ifdef BCLOUD
#ifdef WITH_GPU
#include "paddle/paddle_inference_api.h"
#else
#include "paddle/fluid/inference/api/paddle_inference_api.h"
#endif
#else
#include "paddle_inference_api.h" // NOLINT
#endif
#include <string>
namespace baidu {
namespace paddle_serving {
......
core/general-server/op/general_infer_op.cpp
浏览文件 @ e832b55d
......@@ -60,10 +60,11 @@ int GeneralInferOp::inference() {
const TensorVector *in = &input_blob->tensor_vector;
TensorVector *out = &output_blob->tensor_vector;
int batch_size = input_blob->GetBatchSize();
int batch_size = input_blob->_batch_size;
VLOG(2) << "(logid=" << log_id << ") input batch size: " << batch_size;
output_blob->SetBatchSize(batch_size);
output_blob->_batch_size = batch_size;
VLOG(2) << "(logid=" << log_id << ") infer batch size: " << batch_size;
......
core/general-server/op/general_infer_op.h
浏览文件 @ e832b55d
......@@ -15,17 +15,9 @@
#pragma once
#include <string>
#include <vector>
#ifdef BCLOUD
#ifdef WITH_GPU
#include "paddle/paddle_inference_api.h"
#else
#include "paddle/fluid/inference/api/paddle_inference_api.h"
#endif
#else
#include "paddle_inference_api.h" // NOLINT
#endif
#include "core/general-server/general_model_service.pb.h"
#include "core/general-server/op/general_infer_helper.h"
#include "paddle_inference_api.h" // NOLINT
namespace baidu {
namespace paddle_serving {
......
core/general-server/op/general_reader_op.cpp
浏览文件 @ e832b55d
......@@ -37,9 +37,9 @@ int conf_check(const Request *req,
const std::shared_ptr<PaddleGeneralModelConfig> &model_config) {
int var_num = req->insts(0).tensor_array_size();
if (var_num != model_config->_feed_type.size()) {
VLOG(2) << "var num: " << var_num;
VLOG(2) << "model config var num: " << model_config->_feed_type.size();
LOG(ERROR) << "feed var number not match.";
LOG(ERROR) << "feed var number not match: model config["
<< model_config->_feed_type.size() << "] vs. actual[" << var_num
<< "]";
return -1;
}
......@@ -73,7 +73,6 @@ int GeneralReaderOp::inference() {
// reade request from client
const Request *req = dynamic_cast<const Request *>(get_request_message());
int batch_size = req->insts_size();
int input_var_num = 0;
std::vector<int64_t> elem_type;
std::vector<int64_t> elem_size;
......@@ -82,26 +81,29 @@ int GeneralReaderOp::inference() {
GeneralBlob *res = mutable_data<GeneralBlob>();
TensorVector *out = &res->tensor_vector;
res->SetBatchSize(batch_size);
res->SetLogId(log_id);
if (!res) {
LOG(ERROR) << "Failed get op tls reader object output";
LOG(ERROR) << "(logid=" << log_id
<< ") Failed get op tls reader object output";
}
Timer timeline;
int64_t start = timeline.TimeStampUS();
int var_num = req->insts(0).tensor_array_size();
VLOG(2) << "var num: " << var_num;
VLOG(2) << "(logid=" << log_id << ") var num: " << var_num;
VLOG(2) << "(logid=" << log_id
<< ") start to call load general model_conf op";
VLOG(2) << "start to call load general model_conf op";
baidu::paddle_serving::predictor::Resource &resource =
baidu::paddle_serving::predictor::Resource::instance();
VLOG(2) << "get resource pointer done.";
VLOG(2) << "(logid=" << log_id << ") get resource pointer done.";
std::shared_ptr<PaddleGeneralModelConfig> model_config =
resource.get_general_model_config();
VLOG(2) << "print general model config done.";
VLOG(2) << "(logid=" << log_id << ") print general model config done.";
// TODO(guru4elephant): how to do conditional check?
/*
......@@ -117,17 +119,6 @@ int GeneralReaderOp::inference() {
elem_type.resize(var_num);
elem_size.resize(var_num);
capacity.resize(var_num);
/*
message Tensor {
repeated bytes data = 1;
repeated int32 int_data = 2;
repeated int64 int64_data = 3;
repeated float float_data = 4;
optional int32 elem_type = 5;
repeated int32 shape = 6;
repeated int32 lod = 7; // only for fetch tensor currently
};
*/
// prepare basic information for input
for (int i = 0; i < var_num; ++i) {
paddle::PaddleTensor lod_tensor;
......@@ -143,8 +134,9 @@ message Tensor {
elem_size[i] = sizeof(int32_t);
lod_tensor.dtype = paddle::PaddleDType::INT32;
}
//implement lod tensor here
// implement lod tensor here
if (req->insts(0).tensor_array(i).lod_size() > 0) {
VLOG(2) << "(logid=" << log_id << ") var[" << i << "] is lod_tensor";
lod_tensor.lod.resize(1);
for (int k = 0; k < req->insts(0).tensor_array(i).lod_size(); ++k) {
lod_tensor.lod[0].push_back(req->insts(0).tensor_array(i).lod(k));
......@@ -152,21 +144,24 @@ message Tensor {
capacity[i] = 1;
for (int k = 0; k < req->insts(0).tensor_array(i).shape_size(); ++k) {
int dim = req->insts(0).tensor_array(i).shape(k);
VLOG(2) << "shape for var[" << i << "]: " << dim;
VLOG(2) << "(logid=" << log_id << ") shape for var[" << i
<< "]: " << dim;
capacity[i] *= dim;
lod_tensor.shape.push_back(dim);
}
VLOG(2) << "var[" << i << "] is tensor, capacity: " << capacity[i];
}
else {
VLOG(2) << "(logid=" << log_id << ") var[" << i
<< "] is tensor, capacity: " << capacity[i];
} else {
capacity[i] = 1;
for (int k = 0; k < req->insts(0).tensor_array(i).shape_size(); ++k) {
int dim = req->insts(0).tensor_array(i).shape(k);
VLOG(2) << "shape for var[" << i << "]: " << dim;
VLOG(2) << "(logid=" << log_id << ") shape for var[" << i
<< "]: " << dim;
capacity[i] *= dim;
lod_tensor.shape.push_back(dim);
}
VLOG(2) << "var[" << i << "] is tensor, capacity: " << capacity[i];
VLOG(2) << "(logid=" << log_id << ") var[" << i
<< "] is tensor, capacity: " << capacity[i];
}
lod_tensor.name = model_config->_feed_name[i];
out->push_back(lod_tensor);
......@@ -186,11 +181,12 @@ message Tensor {
} else if (tensor.int_data_size() > 0) {
data_len = tensor.int_data_size();
}
VLOG(2) << "tensor size for var[" << i << "]: " << data_len;
VLOG(2) << "(logid=" << log_id << ") tensor size for var[" << i
<< "]: " << data_len;
tensor_size += data_len;
int cur_len = out->at(i).lod[0].back();
VLOG(2) << "current len: " << cur_len;
VLOG(2) << "(logid=" << log_id << ") current len: " << cur_len;
int sample_len = 0;
if (tensor.shape_size() == 1) {
......@@ -199,21 +195,21 @@ message Tensor {
sample_len = tensor.shape(0);
}
out->at(i).lod[0].push_back(cur_len + sample_len);
VLOG(2) << "new len: " << cur_len + sample_len;
VLOG(2) << "(logid=" << log_id << ") new len: " << cur_len + sample_len;
}
out->at(i).data.Resize(tensor_size * elem_size[i]);
out->at(i).shape = {};
for (int j = 1; j < req->insts(0).tensor_array(i).shape_size(); ++j) {
out->at(i).shape.push_back(req->insts(0).tensor_array(i).shape(j));
}
//if (out->at(i).shape.size() == 1) {
// if (out->at(i).shape.size() == 1) {
// out->at(i).shape.push_back(1);
//}
VLOG(2) << "var[" << i
VLOG(2) << "(logid=" << log_id << ") var[" << i
<< "] is lod_tensor and len=" << out->at(i).lod[0].back();
} else {
out->at(i).data.Resize(batch_size * capacity[i] * elem_size[i]);
VLOG(2) << "var[" << i
VLOG(2) << "(logid=" << log_id << ") var[" << i
<< "] is tensor and capacity=" << batch_size * capacity[i];
}
}
......@@ -222,8 +218,8 @@ message Tensor {
for (int i = 0; i < var_num; ++i) {
if (elem_type[i] == 0) {
int64_t *dst_ptr = static_cast<int64_t *>(out->at(i).data.data());
VLOG(2) << "first element data in var[" << i << "] is "
<< req->insts(0).tensor_array(i).int64_data(0);
VLOG(2) << "(logid=" << log_id << ") first element data in var[" << i
<< "] is " << req->insts(0).tensor_array(i).int64_data(0);
int offset = 0;
for (int j = 0; j < batch_size; ++j) {
int elem_num = req->insts(j).tensor_array(i).int64_data_size();
......@@ -233,8 +229,8 @@ message Tensor {
}
} else if (elem_type[i] == 1) {
float *dst_ptr = static_cast<float *>(out->at(i).data.data());
VLOG(2) << "first element data in var[" << i << "] is "
<< req->insts(0).tensor_array(i).float_data(0);
VLOG(2) << "(logid=" << log_id << ") first element data in var[" << i
<< "] is " << req->insts(0).tensor_array(i).float_data(0);
int offset = 0;
for (int j = 0; j < batch_size; ++j) {
int elem_num = req->insts(j).tensor_array(i).float_data_size();
......@@ -244,8 +240,8 @@ message Tensor {
}
} else if (elem_type[i] == 2) {
int32_t *dst_ptr = static_cast<int32_t *>(out->at(i).data.data());
VLOG(2) << "first element data in var[" << i << "] is "
<< req->insts(0).tensor_array(i).int_data(0);
VLOG(2) << "(logid=" << log_id << ") first element data in var[" << i
<< "] is " << req->insts(0).tensor_array(i).int_data(0);
int offset = 0;
for (int j = 0; j < batch_size; ++j) {
int elem_num = req->insts(j).tensor_array(i).int_data_size();
......@@ -256,15 +252,15 @@ message Tensor {
}
}
VLOG(2) << "output size: " << out->size();
VLOG(2) << "(logid=" << log_id << ") output size: " << out->size();
timeline.Pause();
int64_t end = timeline.TimeStampUS();
res->p_size = 0;
res->_batch_size = batch_size;
AddBlobInfo(res, start);
AddBlobInfo(res, end);
VLOG(2) << "read data from client success";
VLOG(2) << "(logid=" << log_id << ") read data from client success";
return 0;
}
DEFINE_OP(GeneralReaderOp);
......
core/general-server/op/general_reader_op.h
浏览文件 @ e832b55d
......@@ -13,21 +13,13 @@
// limitations under the License.
#pragma once
#include <vector>
#ifdef BCLOUD
#ifdef WITH_GPU
#include "paddle/paddle_inference_api.h"
#else
#include "paddle/fluid/inference/api/paddle_inference_api.h"
#endif
#else
#include "paddle_inference_api.h" // NOLINT
#endif
#include <string>
#include <vector>
#include "core/general-server/general_model_service.pb.h"
#include "core/general-server/load_general_model_service.pb.h"
#include "core/general-server/op/general_infer_helper.h"
#include "core/predictor/framework/resource.h"
#include "paddle_inference_api.h" // NOLINT
namespace baidu {
namespace paddle_serving {
......
core/general-server/op/general_response_op.h
浏览文件 @ e832b55d
......@@ -15,16 +15,8 @@
#pragma once
#include <string>
#include <vector>
#ifdef BCLOUD
#ifdef WITH_GPU
#include "paddle/paddle_inference_api.h"
#else
#include "paddle/fluid/inference/api/paddle_inference_api.h"
#endif
#else
#include "paddle_inference_api.h" // NOLINT
#endif
#include "core/general-server/general_model_service.pb.h"
#include "paddle_inference_api.h" // NOLINT
namespace baidu {
namespace paddle_serving {
......
core/general-server/op/general_text_reader_op.h
浏览文件 @ e832b55d
......@@ -13,21 +13,13 @@
// limitations under the License.
#pragma once
#include <vector>
#ifdef BCLOUD
#ifdef WITH_GPU
#include "paddle/paddle_inference_api.h"
#else
#include "paddle/fluid/inference/api/paddle_inference_api.h"
#endif
#else
#include "paddle_inference_api.h" // NOLINT
#endif
#include <string>
#include <vector>
#include "core/general-server/general_model_service.pb.h"
#include "core/general-server/load_general_model_service.pb.h"
#include "core/general-server/op/general_infer_helper.h"
#include "core/predictor/framework/resource.h"
#include "paddle_inference_api.h" // NOLINT
namespace baidu {
namespace paddle_serving {
......
core/general-server/op/general_text_response_op.h
浏览文件 @ e832b55d
......@@ -15,17 +15,9 @@
#pragma once
#include <string>
#include <vector>
#ifdef BCLOUD
#ifdef WITH_GPU
#include "paddle/paddle_inference_api.h"
#else
#include "paddle/fluid/inference/api/paddle_inference_api.h"
#endif
#else
#include "paddle_inference_api.h" // NOLINT
#endif
#include "core/general-server/general_model_service.pb.h"
#include "core/general-server/op/general_infer_helper.h"
#include "paddle_inference_api.h" // NOLINT
namespace baidu {
namespace paddle_serving {
......
core/predictor/common/inner_common.h
浏览文件 @ e832b55d
......@@ -50,7 +50,7 @@
#include "butil/time.h"
#endif
#include "glog/raw_logging.h"
#define ERROR_STRING_LEN 10240
#include "core/configure/general_model_config.pb.h"
#include "core/configure/include/configure_parser.h"
......
core/predictor/framework/factory.h
浏览文件 @ e832b55d
......@@ -17,7 +17,7 @@
#include <string>
#include <utility>
#include "core/predictor/common/inner_common.h"
#include "glog/raw_logging.h"
namespace baidu {
namespace paddle_serving {
namespace predictor {
......@@ -28,7 +28,12 @@ namespace predictor {
FactoryDerive<D, B>* factory = new (std::nothrow) FactoryDerive<D, B>(); \
if (factory == NULL || \
FactoryPool<B>::instance().register_factory(tag, factory) != 0) { \
RAW_LOG_FATAL("Failed regist factory: %s in macro!", #D); \
char err_str[ERROR_STRING_LEN]; \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Failed regist factory: %s in macro!", \
#D); \
RAW_LOG(FATAL, err_str); \
return -1; \
} \
return 0; \
......@@ -54,7 +59,13 @@ namespace predictor {
if (factory == NULL || \
::baidu::paddle_serving::predictor::FactoryPool<B>::instance() \
.register_factory(#D, factory) != 0) { \
RAW_LOG_FATAL("Failed regist factory: %s->%s in macro!", #D, #B); \
char err_str[ERROR_STRING_LEN]; \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Failed regist factory: %s->%s in macro!", \
#D, \
#B); \
RAW_LOG(FATAL, err_str); \
return; \
} \
return; \
......@@ -66,15 +77,26 @@ namespace predictor {
::baidu::paddle_serving::predictor::FactoryDerive<D, B>* factory = new ( \
::std::nothrow)::baidu::paddle_serving::predictor::FactoryDerive<D, \
B>(); \
char err_str[ERROR_STRING_LEN]; \
if (factory == NULL || \
::baidu::paddle_serving::predictor::FactoryPool<B>::instance() \
.register_factory(N, factory) != 0) { \
RAW_LOG_FATAL( \
"Failed regist factory: %s->%s, tag: %s in macro!", #D, #B, N); \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Failed regist factory: %s->%s, tag: %s in macro!", \
#D, \
#B, \
N); \
RAW_LOG(FATAL, err_str); \
return; \
} \
RAW_LOG_WARNING( \
"Succ regist factory: %s->%s, tag: %s in macro!", #D, #B, N); \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Succ regist factory: %s->%s, tag: %s in macro!", \
#D, \
#B, \
N); \
RAW_LOG(WARNING, err_str); \
return; \
}
......@@ -102,24 +124,35 @@ class FactoryPool {
}
int register_factory(const std::string& tag, FactoryBase<B>* factory) {
char err_str[ERROR_STRING_LEN];
typename std::map<std::string, FactoryBase<B>*>::iterator it =
_pool.find(tag);
if (it != _pool.end()) {
RAW_LOG_FATAL("Insert duplicate with tag: %s", tag.c_str());
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Insert duplicate with tag: %s",
tag.c_str());
RAW_LOG(FATAL, err_str);
return -1;
}
std::pair<typename std::map<std::string, FactoryBase<B>*>::iterator, bool>
r = _pool.insert(std::make_pair(tag, factory));
if (!r.second) {
RAW_LOG_FATAL("Failed insert new factory with: %s", tag.c_str());
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Failed insert new factory with: %s",
tag.c_str());
RAW_LOG(FATAL, err_str);
return -1;
}
RAW_LOG_INFO("Succ insert one factory, tag: %s, base type %s",
tag.c_str(),
typeid(B).name());
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Succ insert one factory, tag: %s, base type %s",
tag.c_str(),
typeid(B).name());
RAW_LOG(INFO, err_str);
return 0;
}
......@@ -127,9 +160,13 @@ class FactoryPool {
typename std::map<std::string, FactoryBase<B>*>::iterator it =
_pool.find(tag);
if (it == _pool.end() || it->second == NULL) {
RAW_LOG_FATAL("Not found factory pool, tag: %s, pool size %u",
tag.c_str(),
_pool.size());
char err_str[ERROR_STRING_LEN];
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Not found factory pool, tag: %s, pool size %u",
tag.c_str(),
_pool.size());
RAW_LOG(FATAL, err_str);
return NULL;
}
......
core/predictor/framework/infer.h
浏览文件 @ e832b55d
......@@ -603,6 +603,7 @@ class VersionedInferEngine : public InferEngine {
LOG(ERROR) << "Failed generate engine with type:" << engine_type;
return -1;
}
#ifndef BCLOUD
VLOG(2) << "FLAGS_logtostderr " << FLAGS_logtostderr;
int tmp = FLAGS_logtostderr;
if (engine->proc_initialize(conf, version) != 0) {
......@@ -611,6 +612,12 @@ class VersionedInferEngine : public InferEngine {
}
VLOG(2) << "FLAGS_logtostderr " << FLAGS_logtostderr;
FLAGS_logtostderr = tmp;
#else
if (engine->proc_initialize(conf, version) != 0) {
LOG(ERROR) << "Failed initialize engine, type:" << engine_type;
return -1;
}
#endif
auto r = _versions.insert(std::make_pair(engine->version(), engine));
if (!r.second) {
LOG(ERROR) << "Failed insert item: " << engine->version()
......
core/predictor/framework/op_repository.h
浏览文件 @ e832b55d
......@@ -62,7 +62,10 @@ class OpRepository {
template <typename OP_TYPE>
void regist_op(std::string op_type) {
_repository[op_type] = &OpFactory<OP_TYPE>::instance();
RAW_LOG_INFO("Succ regist op: %s", op_type.c_str());
char err_str[ERROR_STRING_LEN];
snprintf(
err_str, ERROR_STRING_LEN - 1, "Succ regist op: %s", op_type.c_str());
RAW_LOG(INFO, err_str);
}
Op* get_op(std::string op_type);
......
core/predictor/framework/resource.cpp
浏览文件 @ e832b55d
......@@ -17,6 +17,9 @@
#include <string>
#include "core/predictor/common/inner_common.h"
#include "core/predictor/framework/kv_manager.h"
#ifdef BCLOUD
#include "aipe_sec_client.h" // NOLINT
#endif
namespace baidu {
namespace paddle_serving {
namespace predictor {
......@@ -109,6 +112,42 @@ int Resource::initialize(const std::string& path, const std::string& file) {
}
LOG(WARNING) << "Successfully proc initialized mempool wrapper";
#ifdef WITH_AUTH
std::string product_name_str = resource_conf.auth_product_name();
std::string container_id_str = resource_conf.auth_container_id();
char* product_name = new char[product_name_str.size() + 1];
snprintf(product_name,
product_name_str.size() + 1,
"%s",
product_name_str.c_str());
char* container_id = new char[container_id_str.size() + 1];
snprintf(container_id,
container_id_str.size() + 1,
"%s",
container_id_str.c_str());
aipe_auth_request request;
request.product_name = product_name;
request.container_id = container_id;
request.request_ts = (int64_t)time(NULL);
LOG(INFO) << "\nEasypack info"
<< "\nproduct name: " << request.product_name
<< "\ncontainer_id: " << request.container_id
<< "\nrequest time stamp: " << request.request_ts;
aipe_auth_response response;
response = check_auth(request);
if (response.result == 0) {
LOG(INFO) << "Authentication succeed.";
} else {
LOG(ERROR) << "Authentication failed. Error code: " << response.result;
return -1;
}
#endif
if (FLAGS_enable_model_toolkit) {
int err = 0;
std::string model_toolkit_path = resource_conf.model_toolkit_path();
......
core/predictor/framework/service_manager.h
浏览文件 @ e832b55d
......@@ -23,17 +23,24 @@ namespace predictor {
#define REGIST_FORMAT_SERVICE(svr_name, svr) \
do { \
char err_str[ERROR_STRING_LEN]; \
int ret = \
::baidu::paddle_serving::predictor::FormatServiceManager::instance() \
.regist_service(svr_name, svr); \
if (ret != 0) { \
RAW_LOG_ERROR("Failed regist service[%s][%s]", \
svr_name.c_str(), \
typeid(svr).name()); \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Failed regist service[%s][%s]", \
svr_name.c_str(), \
typeid(svr).name()); \
RAW_LOG(ERROR, err_str); \
} else { \
RAW_LOG_INFO("Success regist service[%s][%s]", \
svr_name.c_str(), \
typeid(svr).name()); \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Success regist service[%s][%s]", \
svr_name.c_str(), \
typeid(svr).name()); \
RAW_LOG(INFO, err_str); \
} \
} while (0)
......@@ -42,31 +49,46 @@ class FormatServiceManager {
typedef google::protobuf::Service Service;
int regist_service(const std::string& svr_name, Service* svr) {
char err_str[ERROR_STRING_LEN];
if (_service_map.find(svr_name) != _service_map.end()) {
RAW_LOG_ERROR("Service[%s][%s] already exist!",
svr_name.c_str(),
typeid(svr).name());
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Service[%s][%s] already exist!",
svr_name.c_str(),
typeid(svr).name());
RAW_LOG(ERROR, err_str);
return -1;
}
std::pair<boost::unordered_map<std::string, Service*>::iterator, bool> ret;
ret = _service_map.insert(std::make_pair(svr_name, svr));
if (ret.second == false) {
RAW_LOG_ERROR("Service[%s][%s] insert failed!",
svr_name.c_str(),
typeid(svr).name());
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Service[%s][%s] insert failed!",
svr_name.c_str(),
typeid(svr).name());
RAW_LOG(ERROR, err_str);
return -1;
}
RAW_LOG_INFO("Service[%s] insert successfully!", svr_name.c_str());
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Service[%s] insert successfully!",
svr_name.c_str());
RAW_LOG(INFO, err_str);
return 0;
}
Service* get_service(const std::string& svr_name) {
char err_str[ERROR_STRING_LEN];
boost::unordered_map<std::string, Service*>::iterator res;
if ((res = _service_map.find(svr_name)) == _service_map.end()) {
RAW_LOG_WARNING("Service[%s] not found in service manager!",
svr_name.c_str());
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Service[%s] not found in service manager!",
svr_name.c_str());
RAW_LOG(WARNING, err_str);
return NULL;
}
return (*res).second;
......
core/predictor/src/pdserving.cpp
浏览文件 @ e832b55d
......@@ -202,8 +202,6 @@ int main(int argc, char** argv) {
}
VLOG(2) << "Succ call pthread worker start function";
#ifndef BCLOUD
if (Resource::instance().general_model_initialize(FLAGS_resource_path,
FLAGS_resource_file) != 0) {
LOG(ERROR) << "Failed to initialize general model conf: "
......@@ -213,6 +211,7 @@ int main(int argc, char** argv) {
VLOG(2) << "Succ initialize general model";
#ifndef BCLOUD
// FATAL messages are output to stderr
FLAGS_stderrthreshold = 3;
#endif
......
core/sdk-cpp/include/abtest.h
浏览文件 @ e832b55d
......@@ -50,9 +50,9 @@ class WeightedRandomRender : public EndpointRouterBase {
Factory<WeightedRandomRender, EndpointRouterBase>* factory =
new (std::nothrow) Factory<WeightedRandomRender, EndpointRouterBase>();
if (factory == NULL) {
RAW_LOG_ERROR(
"Failed regist factory: WeightedRandomRender->EndpointRouterBase in "
"macro!");
RAW_LOG(ERROR,
"Failed regist factory: WeightedRandomRender->EndpointRouterBase \
in macro!");
return -1;
}
......@@ -62,9 +62,9 @@ class WeightedRandomRender : public EndpointRouterBase {
// together.
if (FactoryPool<EndpointRouterBase>::instance().register_factory(
"WeightedRandomRender", factory) != 0) {
RAW_LOG_INFO(
"Factory has been registed: "
"WeightedRandomRender->EndpointRouterBase.");
RAW_LOG(INFO,
"Factory has been registed: \
WeightedRandomRender->EndpointRouterBase.");
}
return 0;
......
core/sdk-cpp/include/factory.h
浏览文件 @ e832b55d
......@@ -18,7 +18,6 @@
#include <utility>
#include "core/sdk-cpp/include/common.h"
#include "core/sdk-cpp/include/stub_impl.h"
#include "glog/raw_logging.h"
namespace baidu {
namespace paddle_serving {
......@@ -28,12 +27,20 @@ namespace sdk_cpp {
namespace brpc = baidu::rpc;
#endif
#define ERROR_STRING_LEN 10240
#define INLINE_REGIST_OBJECT(D, B, E) \
do { \
Factory<D, B>* factory = new (std::nothrow) Factory<D, B>(); \
if (factory == NULL || \
FactoryPool<B>::instance().register_factory(#D, factory) != 0) { \
RAW_LOG_ERROR("Failed regist factory: %s->%s in macro!", #D, #B); \
char err_str[ERROR_STRING_LEN]; \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Failed regist factory: %s->%s in macro!", \
#D, \
#B); \
RAW_LOG(ERROR, err_str); \
return E; \
} \
} while (0)
......@@ -43,7 +50,12 @@ namespace brpc = baidu::rpc;
Factory<D, B>* factory = new (std::nothrow) Factory<D, B>(); \
if (factory == NULL || \
FactoryPool<B>::instance().register_factory(tag, factory) != 0) { \
RAW_LOG_ERROR("Failed regist factory: %s in macro!", #D); \
char err_str[ERROR_STRING_LEN]; \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Failed regist factory: %s in macro!", \
#D); \
RAW_LOG(ERROR, err_str); \
return -1; \
} \
return 0; \
......@@ -66,7 +78,13 @@ namespace brpc = baidu::rpc;
if (factory == NULL || \
::baidu::paddle_serving::sdk_cpp::FactoryPool<B>::instance() \
.register_factory(#D, factory) != 0) { \
RAW_LOG_ERROR("Failed regist factory: %s->%s in macro!", #D, #B); \
char err_str[ERROR_STRING_LEN]; \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Failed regist factory: %s->%s in macro!", \
#D, \
#B); \
RAW_LOG(ERROR, err_str); \
return; \
} \
return; \
......@@ -80,8 +98,14 @@ namespace brpc = baidu::rpc;
if (factory == NULL || \
::baidu::paddle_serving::sdk_cpp::FactoryPool<B>::instance() \
.register_factory(T, factory) != 0) { \
RAW_LOG_ERROR( \
"Failed regist factory: %s->%s, tag %s in macro!", #D, #B, T); \
char err_str[ERROR_STRING_LEN]; \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Failed regist factory: %s->%s, tag %s in macro!", \
#D, \
#B, \
T); \
RAW_LOG(ERROR, err_str); \
return; \
} \
return; \
......@@ -108,8 +132,13 @@ namespace brpc = baidu::rpc;
::baidu::paddle_serving::sdk_cpp::FactoryPool< \
::baidu::paddle_serving::sdk_cpp::Stub>::instance() \
.register_factory(T, factory) != 0) { \
RAW_LOG_ERROR( \
"Failed regist factory: %s->Stub, tag: %s in macro!", #D, T); \
char err_str[ERROR_STRING_LEN]; \
snprintf(err_str, \
ERROR_STRING_LEN - 1, \
"Failed regist factory: %s->Stub, tag: %s in macro!", \
#D, \
T); \
RAW_LOG(ERROR, err_str); \
return; \
} \
return; \
......@@ -146,14 +175,24 @@ class FactoryPool {
typename std::map<std::string, FactoryBase<B>*>::iterator it =
_pool.find(tag);
if (it != _pool.end()) {
RAW_LOG_ERROR("Insert duplicate with tag: %s", tag.c_str());
char err_str[ERROR_STRING_LEN];
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Insert duplicate with tag: %s",
tag.c_str());
RAW_LOG(ERROR, err_str);
return -1;
}
std::pair<typename std::map<std::string, FactoryBase<B>*>::iterator, bool>
r = _pool.insert(std::make_pair(tag, factory));
if (!r.second) {
RAW_LOG_ERROR("Failed insert new factory with: %s", tag.c_str());
char err_str[ERROR_STRING_LEN];
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Failed insert new factory with: %s",
tag.c_str());
RAW_LOG(ERROR, err_str);
return -1;
}
......@@ -164,9 +203,13 @@ class FactoryPool {
typename std::map<std::string, FactoryBase<B>*>::iterator it =
_pool.find(tag);
if (it == _pool.end() || it->second == NULL) {
RAW_LOG_ERROR("Not found factory pool, tag: %s, pool size: %u",
tag.c_str(),
_pool.size());
char err_str[ERROR_STRING_LEN];
snprintf(err_str,
ERROR_STRING_LEN - 1,
"Not found factory pool, tag: %s, pool size: %u",
tag.c_str(),
_pool.size());
RAW_LOG(ERROR, err_str);
return NULL;
}
......
doc/PIPELINE_OP.md
浏览文件 @ e832b55d
......@@ -90,5 +90,3 @@ rec_op = RecOp(
| client_config | 客户端配置文件 | Op接收请求作为客户端访问rpc服务，因此需要客户端的配置文件 |
| local predictor限定 | | |
| model_config | 模型配置文件 | 由于local predictor和Op运行在一台机器上，因此需要模型配置来启动local predictor |
paddle_inference/inferencer-fluid-cpu/include/fluid_cpu_engine.h
浏览文件 @ e832b55d
......@@ -23,6 +23,7 @@
#include "core/configure/inferencer_configure.pb.h"
#include "core/predictor/framework/infer.h"
#include "paddle_inference_api.h" // NOLINT
//#include "predictor/framework/infer.h"
namespace baidu {
namespace paddle_serving {
......
python/paddle_serving_app/local_predict.py
浏览文件 @ e832b55d
......@@ -121,7 +121,7 @@ class Debugger(object):
name])
if self.feed_types_[name] == 0:
feed[name] = feed[name].astype("int64")
elif self.feed_types_[name] == 1:
elif self.feed_types_[name] == 1:
feed[name] = feed[name].astype("float32")
elif self.feed_types_[name] == 2:
feed[name] = feed[name].astype("int32")
......@@ -131,7 +131,7 @@ class Debugger(object):
if "{}.lod".format(name) in feed:
input_tensor.set_lod([feed["{}.lod".format(name)]])
if batch == False:
input_tensor.copy_from_cpu(feed[name][np.newaxis,:])
input_tensor.copy_from_cpu(feed[name][np.newaxis, :])
else:
input_tensor.copy_from_cpu(feed[name])
output_tensors = []
......
python/paddle_serving_client/__init__.py
浏览文件 @ e832b55d
......@@ -233,7 +233,12 @@ class Client(object):
# key))
pass
def predict(self, feed=None, fetch=None, batch=False, need_variant_tag=False, log_id=0):
def predict(self,
feed=None,
fetch=None,
batch=False,
need_variant_tag=False,
log_id=0):
self.profile_.record('py_prepro_0')
if feed is None or fetch is None:
......@@ -262,7 +267,7 @@ class Client(object):
int_shape = []
lod_slot_batch = []
float_shape = []
fetch_names = []
counter = 0
batch_size = len(feed_batch)
......@@ -292,7 +297,7 @@ class Client(object):
int_feed_names.append(key)
shape_lst = []
if batch == False:
feed_i[key] = feed_i[key][np.newaxis,:]
feed_i[key] = feed_i[key][np.newaxis, :]
shape_lst.append(1)
if isinstance(feed_i[key], np.ndarray):
print("feed_i_key shape", feed_i[key].shape)
......@@ -318,7 +323,7 @@ class Client(object):
float_feed_names.append(key)
shape_lst = []
if batch == False:
feed_i[key] = feed_i[key][np.newaxis,:]
feed_i[key] = feed_i[key][np.newaxis, :]
shape_lst.append(1)
if isinstance(feed_i[key], np.ndarray):
print("feed_i_key shape", feed_i[key].shape)
......@@ -349,10 +354,11 @@ class Client(object):
if self.all_numpy_input:
res = self.client_handle_.numpy_predict(
float_slot_batch, float_feed_names, float_shape, int_slot_batch,
int_feed_names, int_shape, lod_slot_batch, fetch_names, result_batch_handle,
self.pid, log_id)
int_feed_names, int_shape, lod_slot_batch, fetch_names,
result_batch_handle, self.pid, log_id)
elif self.has_numpy_input == False:
raise ValueError("Please make sure all of your inputs are numpy array")
raise ValueError(
"Please make sure all of your inputs are numpy array")
else:
raise ValueError(
"Please make sure the inputs are all in list type or all in numpy.array type"
......
python/paddle_serving_server/__init__.py
浏览文件 @ e832b55d
......@@ -103,8 +103,8 @@ class OpSeqMaker(object):
elif len(node.dependencies) == 1:
if node.dependencies[0].name != self.workflow.nodes[-1].name:
raise Exception(
'You must add op in order in OpSeqMaker. The previous op is {}, but the current op is followed by {}.'.
format(node.dependencies[0].name, self.workflow.nodes[
'You must add op in order in OpSeqMaker. The previous op is {}, but the current op is followed by {}.'
.format(node.dependencies[0].name, self.workflow.nodes[
-1].name))
self.workflow.nodes.extend([node])
......@@ -157,6 +157,8 @@ class Server(object):
self.cur_path = os.getcwd()
self.use_local_bin = False
self.mkl_flag = False
self.product_name = None
self.container_id = None
self.model_config_paths = None # for multi-model in a workflow
def set_max_concurrency(self, concurrency):
......@@ -191,6 +193,16 @@ class Server(object):
def set_ir_optimize(self, flag=False):
self.ir_optimization = flag
def set_product_name(self, product_name=None):
if product_name == None:
raise ValueError("product_name can't be None.")
self.product_name = product_name
def set_container_id(self, container_id):
if container_id == None:
raise ValueError("container_id can't be None.")
self.container_id = container_id
def check_local_bin(self):
if "SERVING_BIN" in os.environ:
self.use_local_bin = True
......@@ -254,6 +266,10 @@ class Server(object):
self.resource_conf.model_toolkit_file = self.model_toolkit_fn
self.resource_conf.general_model_path = workdir
self.resource_conf.general_model_file = self.general_model_config_fn
if self.product_name != None:
self.resource_conf.auth_product_name = self.product_name
if self.container_id != None:
self.resource_conf.auth_container_id = self.container_id
def _write_pb_str(self, filepath, pb_obj):
with open(filepath, "w") as fout:
......@@ -351,8 +367,8 @@ class Server(object):
if os.path.exists(tar_name):
os.remove(tar_name)
raise SystemExit(
'Download failed, please check your network or permission of {}.'.
format(self.module_path))
'Download failed, please check your network or permission of {}.'
.format(self.module_path))
else:
try:
print('Decompressing files ..')
......@@ -363,8 +379,8 @@ class Server(object):
if os.path.exists(exe_path):
os.remove(exe_path)
raise SystemExit(
'Decompressing failed, please check your permission of {} or disk space left.'.
format(self.module_path))
'Decompressing failed, please check your permission of {} or disk space left.'
.format(self.module_path))
finally:
os.remove(tar_name)
#release lock
......@@ -541,7 +557,6 @@ class MultiLangServerServiceServicer(multi_lang_general_model_service_pb2_grpc.
results, tag = ret
resp.tag = tag
resp.err_code = 0
if not self.is_multi_model_:
results = {'general_infer_0': results}
for model_name, model_result in results.items():
......
python/paddle_serving_server/serve.py
浏览文件 @ e832b55d
......@@ -58,6 +58,16 @@ def parse_args(): # pylint: disable=doc-string-missing
default=False,
action="store_true",
help="Use Multi-language-service")
parser.add_argument(
"--product_name",
type=str,
default=None,
help="product_name for authentication")
parser.add_argument(
"--container_id",
type=str,
default=None,
help="container_id for authentication")
return parser.parse_args()
......@@ -101,6 +111,10 @@ def start_standard_model(): # pylint: disable=doc-string-missing
server.use_mkl(use_mkl)
server.set_max_body_size(max_body_size)
server.set_port(port)
if args.product_name != None:
server.set_product_name(args.product_name)
if args.container_id != None:
server.set_container_id(args.container_id)
server.load_model_config(model)
server.prepare_server(workdir=workdir, port=port, device=device)
......
python/paddle_serving_server_gpu/__init__.py
浏览文件 @ e832b55d
......@@ -73,6 +73,16 @@ def serve_args():
default=False,
action="store_true",
help="Use Multi-language-service")
parser.add_argument(
"--product_name",
type=str,
default=None,
help="product_name for authentication")
parser.add_argument(
"--container_id",
type=str,
default=None,
help="container_id for authentication")
return parser.parse_args()
......@@ -141,8 +151,8 @@ class OpSeqMaker(object):
elif len(node.dependencies) == 1:
if node.dependencies[0].name != self.workflow.nodes[-1].name:
raise Exception(
'You must add op in order in OpSeqMaker. The previous op is {}, but the current op is followed by {}.'.
format(node.dependencies[0].name, self.workflow.nodes[
'You must add op in order in OpSeqMaker. The previous op is {}, but the current op is followed by {}.'
.format(node.dependencies[0].name, self.workflow.nodes[
-1].name))
self.workflow.nodes.extend([node])
......@@ -196,6 +206,8 @@ class Server(object):
self.use_local_bin = False
self.gpuid = 0
self.model_config_paths = None # for multi-model in a workflow
self.product_name = None
self.container_id = None
def set_max_concurrency(self, concurrency):
self.max_concurrency = concurrency
......@@ -229,6 +241,16 @@ class Server(object):
def set_ir_optimize(self, flag=False):
self.ir_optimization = flag
def set_product_name(self, product_name=None):
if product_name == None:
raise ValueError("product_name can't be None.")
self.product_name = product_name
def set_container_id(self, container_id):
if container_id == None:
raise ValueError("container_id can't be None.")
self.container_id = container_id
def check_local_bin(self):
if "SERVING_BIN" in os.environ:
self.use_local_bin = True
......@@ -302,6 +324,10 @@ class Server(object):
self.resource_conf.model_toolkit_file = self.model_toolkit_fn
self.resource_conf.general_model_path = workdir
self.resource_conf.general_model_file = self.general_model_config_fn
if self.product_name != None:
self.resource_conf.auth_product_name = self.product_name
if self.container_id != None:
self.resource_conf.auth_container_id = self.container_id
def _write_pb_str(self, filepath, pb_obj):
with open(filepath, "w") as fout:
......@@ -393,8 +419,8 @@ class Server(object):
if os.path.exists(tar_name):
os.remove(tar_name)
raise SystemExit(
'Download failed, please check your network or permission of {}.'.
format(self.module_path))
'Download failed, please check your network or permission of {}.'
.format(self.module_path))
else:
try:
print('Decompressing files ..')
......@@ -405,8 +431,8 @@ class Server(object):
if os.path.exists(exe_path):
os.remove(exe_path)
raise SystemExit(
'Decompressing failed, please check your permission of {} or disk space left.'.
format(self.module_path))
'Decompressing failed, please check your permission of {} or disk space left.'
.format(self.module_path))
finally:
os.remove(tar_name)
#release lock
......
python/paddle_serving_server_gpu/serve.py
浏览文件 @ e832b55d
......@@ -65,6 +65,11 @@ def start_gpu_card_model(index, gpuid, args): # pylint: disable=doc-string-miss
server.set_ir_optimize(ir_optim)
server.set_max_body_size(max_body_size)
if args.product_name != None:
server.set_product_name(args.product_name)
if args.container_id != None:
server.set_container_id(args.container_id)
server.load_model_config(model)
server.prepare_server(workdir=workdir, port=port, device=device)
if gpuid >= 0:
......@@ -83,8 +88,8 @@ def start_multi_card(args): # pylint: disable=doc-string-missing
for ids in gpus:
if int(ids) >= len(env_gpus):
print(
" Max index of gpu_ids out of range, the number of CUDA_VISIBLE_DEVICES is {}.".
format(len(env_gpus)))
" Max index of gpu_ids out of range, the number of CUDA_VISIBLE_DEVICES is {}."
.format(len(env_gpus)))
exit(-1)
else:
env_gpus = []
......
python/setup.py.app.in
浏览文件 @ e832b55d
......@@ -43,7 +43,7 @@ if '${PACK}' == 'ON':
REQUIRED_PACKAGES = [
'six >= 1.10.0', 'sentencepiece', 'opencv-python<=4.2.0.32', 'pillow',
'shapely', 'pyclipper'
'shapely<=1.6.1', 'pyclipper'
]
packages=['paddle_serving_app',
......
Markdown is supported
0% 或 .
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 或