moving towards a tflow example

This commit is contained in:
Kevin Darby committed 2019-01-07 17:54:39 -05:00
1 parent 7c42e75ddd
commit 59369e1b2b
13 files changed
+171 -187

No files matched your search

+40
View File
@@ -108,6 +108,46 @@
"/home/bts/spark/": "${workspaceRoot}"
}
},
{
"name": "TFTest ",
"type": "cppdbg",
"request": "launch",
"program": "/home/bts/spark/build/ninja_debug/sdk/examples/tf_test",
"args": [
],
"stopAtEntry": false,
"cwd": "/home/bts/spark/build/ninja_debug/sdk/examples/",
"environment": [],
"externalConsole": true,
"MIMode": "gdb",
"preLaunchTask": "",
"logging": {
"engineLogging": false
},
"setupCommands": [
{
"description": "Enable pretty-printing for gdb",
"text": "enable pretty-printer",
"ignoreFailures": true
},
{
"description": "Skip STL",
"text": "skip -gfi /usr/include/c++/7/bits/*.h",
"ignoreFailures": true
}
],
"pipeTransport": {
"pipeCwd": "${workspaceRoot}",
"pipeProgram": "c:\\program files\\git\\usr\\bin\\ssh.exe",
"pipeArgs": [
"bts@sparkvm"
],
"debuggerPath": "/usr/bin/gdb"
},
"sourceFileMap": {
"/home/bts/spark/": "${workspaceRoot}"
}
},
{
"name": "Spreader Tests",
"type": "cppdbg",
+4 -1
View File
@@ -34,7 +34,7 @@ if (WIN32)
else ()
# so that sparkd looks for libspark.so in its directory
set (CMAKE_BUILD_WITH_INSTALL_RPATH TRUE)
set (CMAKE_INSTALL_RPATH ".")
set (CMAKE_INSTALL_RPATH ".:./lib")
set (CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} \
-fPIC -Wall -pedantic -Wextra -Werror \
@@ -58,6 +58,9 @@ else ()
COMMAND mkdir ARGS -p ${CMAKE_BINARY_DIR}/server/plugins
COMMAND ln ARGS -sf ${CMAKE_BINARY_DIR}/sdk/examples/order_plugin_example.so ${CMAKE_BINARY_DIR}/server/plugins/
COMMAND ln ARGS -sf ${CMAKE_BINARY_DIR}/sdk/examples/ai_order_plugin_example.so ${CMAKE_BINARY_DIR}/server/plugins/
COMMAND mkdir -p ${CMAKE_BINARY_DIR}/server/lib
COMMAND ln ARGS -sf ${EXTERNALS_ROOT}/lib/libtensorflow.so ${CMAKE_BINARY_DIR}/server/lib/
COMMAND ln ARGS -sf ${EXTERNALS_ROOT}/lib/libtensorflow_framework.so ${CMAKE_BINARY_DIR}/server/lib/
COMMENT "Creating server runtime layout"
DEPENDS sparkd data_files_dir)
+3 -2
View File
@@ -55,15 +55,16 @@ g++ -isystem ../externals/include \
tf_test.cpp \
../tensorflow/tfexecute.cpp \
../tensorflow/tfloader.cpp \
../tensorflow/tfstatus.cpp \
-Wl,-rpath,. \
-Wl,-rpath,.. \
-Wl,-rpath,../lib \
-L ../lib \
-l spark \
-L ../externals/lib \
-l tensorflow \
-l tensorflow_framework \
-l tensorflow-core \
-l nsync \
-lpthread \
-llttng-ust \
-ldl
+10 -10
View File
@@ -1,18 +1,18 @@
#include "tensorflow/core/platform/env.h"
#include "tensorflow/core/public/session.h"
#include "../tensorflow/tfstatus.h"
#include "../tensorflow/tfloader.h"
#include <iostream>
using namespace std;
using namespace tensorflow;
using namespace tensorflow::c;
int main()
{
Session* session;
Status status = NewSession(SessionOptions(), &session);
if (!status.ok()) {
cout << status.ToString() << "\n";
return 1;
}
cout << "Session successfully created.\n";
TF_Session* session = nullptr;
TFStatus status = bts::ml::create_tf_session(session);
if (!status.ok()) {
cout << status.code << "\n";
return 1;
}
cout << "Session successfully created.\n";
}
@@ -1,93 +0,0 @@
/* Copyright 2016 Google Inc. All Rights Reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/
#ifndef TENSORFLOW_SERVING_SERVABLES_TENSORFLOW_SAVED_MODEL_BUNDLE_FACTORY_H_
#define TENSORFLOW_SERVING_SERVABLES_TENSORFLOW_SAVED_MODEL_BUNDLE_FACTORY_H_
#include "tensorflow/cc/saved_model/loader.h"
#include "tensorflow/core/lib/core/status.h"
#include "tensorflow/core/platform/macros.h"
#include "tensorflow_serving/servables/tensorflow/session_bundle_config.pb.h"
namespace tensorflow {
namespace serving {
/// A factory that creates SavedModelBundles from SavedModel or SessionBundle
/// export paths.
///
/// The emitted sessions only support Run(), and although not enforced it is
/// expected that the client will only make non-mutating Run() calls. (If this
/// restriction, which we've added as a safety measure, is problematic for your
/// use-case please contact the TensorFlow Serving team to discuss disabling
/// it.)
///
/// If the config calls for batching, the emitted sessions automatically batch
/// Run() calls behind the scenes, using a SharedBatchScheduler owned by the
/// factory. The 'config.num_batch_threads' threads are shared across all
/// session instances created by this factory. However, each session has its own
/// dedicated queue of size 'config.max_enqueued_batches'.
///
/// The factory can also estimate the resource (e.g. RAM) requirements of a
/// SavedModelBundle based on the SavedModel (i.e. prior to loading the
/// session).
///
/// This class is thread-safe.
class SavedModelBundleFactory {
public:
/// Instantiates a SavedModelBundleFactory using a config.
///
/// @param config Config with initialization options.
/// @param factory Newly created factory if the returned Status is OK.
static Status Create(const SessionBundleConfig& config,
std::unique_ptr<SavedModelBundleFactory>* factory);
/// Instantiates a bundle from a given export or SavedModel path.
///
/// @param path Path to the model.
/// @param bundle Newly created SavedModelBundle if the returned Status is
/// OK.
Status CreateSavedModelBundle(const string& path,
std::unique_ptr<SavedModelBundle>* bundle);
/// Estimates the resources a SavedModel bundle will use once loaded, from its
/// export path.
///
/// @param path Path to the model.
/// @param estimate Output resource usage estimates. Different kinds of
/// resources (e.g. CPU, RAM, etc.) may get populated.
// Status EstimateResourceRequirement(const string& path,
// ResourceAllocation* estimate) const;
// const SessionBundleConfig& config() const { return config_; }
private:
// using Batcher = SharedBatchScheduler<BatchingSessionTask>;
// SavedModelBundleFactory(const SessionBundleConfig& config,
// std::shared_ptr<Batcher> batch_scheduler);
// const SessionBundleConfig config_;
// // A shared batch scheduler. One queue is used for each session this factory
// // emits. If batching is not configured, this remains null.
// std::shared_ptr<Batcher> batch_scheduler_;
TF_DISALLOW_COPY_AND_ASSIGN(SavedModelBundleFactory);
};
} // namespace serving
} // namespace tensorflow
#endif // TENSORFLOW_SERVING_SERVABLES_TENSORFLOW_SAVED_MODEL_BUNDLE_FACTORY_H_
+10
View File
@@ -0,0 +1,10 @@
#pragma once
namespace tensorflow
{
namespace c
{
#include "tensorflow/c/c_api.h"
}
}
+11 -14
View File
@@ -1,8 +1,5 @@
#include "tfexecute.h"
#include "tensorflow/core/public/session.h"
#include "tensorflow/cc/saved_model/loader.h"
using namespace tensorflow;
using std::vector;
@@ -11,17 +8,17 @@ namespace bts
namespace ml
{
tensorflow::Status execute(const tensorflow::SavedModelBundle& bundle,
const vector<std::pair<std::string, tensorflow::Tensor> >& inputs,
const vector<std::string>& output_tensor_names,
const vector<std::string>& target_node_names,
vector<tensorflow::Tensor>* outputs)
{
tensorflow::RunMetadata* run_metadata = nullptr;
tensorflow::RunOptions run_options;
Status status = bundle.session->Run(run_options, inputs, output_tensor_names, target_node_names, outputs, run_metadata);
return status;
}
// tensorflow::Status execute(const tensorflow::SavedModelBundle& bundle,
// const vector<std::pair<std::string, tensorflow::Tensor> >& inputs,
// const vector<std::string>& output_tensor_names,
// const vector<std::string>& target_node_names,
// vector<tensorflow::Tensor>* outputs)
// {
// tensorflow::RunMetadata* run_metadata = nullptr;
// tensorflow::RunOptions run_options;
// Status status = bundle.session->Run(run_options, inputs, output_tensor_names, target_node_names, outputs, run_metadata);
// return status;
// }
}
+7 -6
View File
@@ -1,8 +1,9 @@
#pragma once
#include "tfstatus.h"
#include <vector>
#include <string>
#include "tensorflow/core/platform/env.h"
namespace tensorflow
{
@@ -15,10 +16,10 @@ namespace bts
namespace ml
{
tensorflow::Status execute(const tensorflow::SavedModelBundle& bundle,
const std::vector<std::pair<std::string, tensorflow::Tensor> >& inputs,
const std::vector<std::string>& output_tensor_names,
const std::vector<std::string>& target_node_names,
std::vector<tensorflow::Tensor>* outputs);
// tensorflow::c::TFStatus execute(const tensorflow::c::TF_Session& session,
// const std::vector<std::pair<std::string, tensorflow::c::TF_Tensor> >& inputs,
// const std::vector<std::string>& output_tensor_names,
// const std::vector<std::string>& target_node_names,
// std::vector<tensorflow::Tensor>* outputs);
}
}
+40 -35
View File
@@ -1,12 +1,10 @@
#include "tfloader.h"
#include "lib/utils/helper_macros.h"
#include <vector>
//no c++ yet
//no c++ api yet
//https://github.com/tensorflow/tensorflow/blob/master/tensorflow/examples/label_image/main.cc
#include "tensorflow/cc/saved_model/loader.h"
using namespace tensorflow;
using namespace tensorflow::c;
using std::string;
using std::vector;
@@ -15,7 +13,44 @@ namespace bts
{
namespace ml
{
tensorflow::c::TFStatus create_tf_session(TF_Session*& session)
{
TF_Graph* graph = TF_NewGraph();
TF_SessionOptions* options = TF_NewSessionOptions();
TF_Status* status = TF_NewStatus();
session = TF_NewSession(graph, options, status);
TFStatus stat(*status);
TF_DeleteStatus(status);
return stat;
}
tensorflow::c::TFStatus load_tf_model(const std::string& export_dir, const std::unordered_set<std::string>& tags,
TF_Session*& session)
{
vector<char*> ctags;
for(auto& t : tags)
ctags.push_back(const_cast<char*>(t.c_str()));
TF_SessionOptions* session_options = TF_NewSessionOptions();
TF_Buffer* run_options = nullptr;
TF_Buffer* meta_graph_def = TF_NewBuffer(); //TODO get the protos here
TF_Graph* graph = TF_NewGraph();
TF_Status* status = TF_NewStatus();
session = TF_LoadSessionFromSavedModel(session_options, run_options, export_dir.c_str(), &ctags[0], ctags.size(),
graph, meta_graph_def, status);
TFStatus stat(*status);
TF_DeleteSessionOptions(session_options);
TF_DeleteBuffer(run_options);
TF_DeleteBuffer(meta_graph_def);
TF_DeleteStatus(status);
return stat;
}
}
}
// tensorflow::Status load_tf_model(const std::string& export_dir, const std::unordered_set<string>& tags, SavedModelBundle* const bundle)
// {
// tensorflow::SessionOptions session_options; //defaults
@@ -42,34 +77,4 @@ namespace ml
// stat = factory->CreateSavedModelBundle(export_dir, bundle);
// }
// return stat;
// }
tensorflow::c::TFStatus load_tf_model(const std::string& export_dir, const std::unordered_set<std::string>& tags,
TF_Session** session)
{
vector<char*> ctags;
for(auto& t : tags)
ctags.push_back(const_cast<char*>(t.c_str()));
TF_SessionOptions* session_options = TF_NewSessionOptions();
TF_Buffer* run_options = nullptr;
TF_Buffer* meta_graph_def = TF_NewBuffer(); //TODO get the protos here
TF_Graph* graph = TF_NewGraph();
TF_Status* status = TF_NewStatus();
*session = TF_LoadSessionFromSavedModel(session_options, run_options, export_dir.c_str(), &ctags[0], ctags.size(),
graph, meta_graph_def, status);
std::string msg(TF_Message(status));
TF_Code code = TF_GetCode(status);
TFStatus stat;
stat.error = msg;
stat.code = code;
TF_DeleteSessionOptions(session_options);
TF_DeleteBuffer(run_options);
TF_DeleteBuffer(meta_graph_def);
TF_DeleteStatus(status);
return stat;
}
}
}
// }
+4 -25
View File
@@ -1,38 +1,17 @@
#pragma once
#include "tensorflow/core/platform/env.h"
#include "tfstatus.h"
#include <unordered_set>
#include <string>
#include <memory>
namespace tensorflow
{
struct SavedModelBundle;
namespace c
{
#include "tensorflow/c/c_api.h"
struct TFStatus
{
std::string error;
TF_Code code;
bool ok() { return (int)code == 0; }
};
}
}
namespace bts
{
namespace ml
{
///call Status.ok() for a boolean
// tensorflow::Status load_tf_model(const std::string& export_dir, const std::unordered_set<std::string>& tags,
// tensorflow::SavedModelBundle* const bundle);
// tensorflow::Status load_tf_model(const std::string& export_dir, const std::unordered_set<std::string>& tags,
// std::unique_ptr<tensorflow::SavedModelBundle>* bundle);
tensorflow::c::TFStatus create_tf_session(tensorflow::c::TF_Session*& session);
tensorflow::c::TFStatus load_tf_model(const std::string& export_dir, const std::unordered_set<std::string>& tags,
tensorflow::c::TF_Session** session);
tensorflow::c::TF_Session*& session);
}
}
+21
View File
@@ -0,0 +1,21 @@
#include "tfstatus.h"
namespace tensorflow
{
namespace c
{
TFStatus::TFStatus(TF_Status& stat)
{
std::string msg(TF_Message(&stat));
this->error = msg;
this->code = TF_GetCode(&stat);
}
bool TFStatus::ok() const
{
return (int)code == 0;
}
}}
+20
View File
@@ -0,0 +1,20 @@
#pragma once
#include <string>
#include "tfapi.h"
namespace tensorflow
{
namespace c
{
struct TFStatus
{
TFStatus(TF_Status&);
bool ok() const;
std::string error;
TF_Code code;
};
}
}