Skip to content
GitLab
Commit 5e622c07 authored by Dmitry I. Lyakh's avatar Dmitry I. Lyakh
Browse files

Fixed ExaTN tests 



Signed-off-by: default avatarDmitry I. Lyakh <quant4me@gmail.com>
parent ed7d229b
Pipeline #165209 failed with stage
in 4 minutes and 12 seconds
Hide whitespace changes
Inline Side-by-side
src/exatn/num_server.cpp
View file @ 5e622c07
/** ExaTN::Numerics: Numerical server
REVISION: 2021/08/21
REVISION: 2021/09/21
Copyright (C) 2018-2021 Dmitry I. Lyakh (Liakh)
Copyright (C) 2018-2021 Oak Ridge National Laboratory (UT-Battelle) **/
......@@ -1433,14 +1433,14 @@ bool NumServer::initTensorsRnd(TensorNetwork & tensor_network)
for(auto tens = tensor_network.begin(); tens != tensor_network.end(); ++tens){
auto tensor = tens->second.getTensor();
const auto & tens_name = tensor->getName();
if(tensorAllocated(tens_name)){
if(tens->first == 0){ //output tensor
success = initTensor(tens_name,0.0);
}else{ //input tensor
if(tens->first != 0){ //input tensor
if(tensorAllocated(tens_name)){
success = initTensorRnd(tens_name);
}else{
success = false;
}
}else{
success = false;
}else{ //output tensor
if(tensorAllocated(tens_name)) success = initTensor(tens_name,0.0);
}
if(!success) break;
}
......@@ -1453,14 +1453,14 @@ bool NumServer::initTensorsRndSync(TensorNetwork & tensor_network)
for(auto tens = tensor_network.begin(); tens != tensor_network.end(); ++tens){
auto tensor = tens->second.getTensor();
const auto & tens_name = tensor->getName();
if(tensorAllocated(tens_name)){
if(tens->first == 0){ //output tensor
success = initTensorSync(tens_name,0.0);
}else{ //input tensor
if(tens->first != 0){ //input tensor
if(tensorAllocated(tens_name)){
success = initTensorRndSync(tens_name);
}else{
success = false;
}
}else{
success = false;
}else{ //output tensor
if(tensorAllocated(tens_name)) success = initTensorSync(tens_name,0.0);
}
if(!success) break;
}
......
src/exatn/num_server.hpp
View file @ 5e622c07
/** ExaTN::Numerics: Numerical server
REVISION: 2021/08/20
REVISION: 2021/09/21
Copyright (C) 2018-2021 Dmitry I. Lyakh (Liakh)
Copyright (C) 2018-2021 Oak Ridge National Laboratory (UT-Battelle) **/
......
src/exatn/tests/NumServerTester.cpp
View file @ 5e622c07
......@@ -16,7 +16,7 @@
#include "errors.hpp"
//Test activation:
/*#define EXATN_TEST0
#define EXATN_TEST0
#define EXATN_TEST1
#define EXATN_TEST2
#define EXATN_TEST3
......@@ -41,10 +41,10 @@
#define EXATN_TEST22
#define EXATN_TEST23
#define EXATN_TEST24
#define EXATN_TEST25*/
#define EXATN_TEST25
#define EXATN_TEST26 //requires input file from source
//#define EXATN_TEST27
//#define EXATN_TEST28
#define EXATN_TEST27
#define EXATN_TEST28
#ifdef EXATN_TEST0
......@@ -1961,9 +1961,7 @@ TEST(NumServerTester, neurIPS) {
const auto TENS_ELEM_TYPE = TensorElementType::COMPLEX32;
//exatn::resetLoggingLevel(1,2); //debug
//exatn::activateContrSeqCaching(true);
exatn::activateContrSeqCaching(true);
//exatn::activateFastMath();
bool success = true;
......@@ -1991,16 +1989,19 @@ TEST(NumServerTester, neurIPS) {
success = exatn::initTensorRnd("E"); assert(success);
success = exatn::initTensorRnd("F"); assert(success);
success = exatn::initTensorRnd("G"); assert(success);
success = exatn::sync(); assert(success);
auto flops = exatn::getTotalFlopCount();
auto time_start = exatn::Timer::timeInSecHR();
success = exatn::evaluateTensorNetwork("MERA1d","Z(z0,z1,z2,z3)+=A(a0,a1,a2,z2)*B(b0,b1,b2,z3)*C(c0,c1,a2,b0)*D(d0,d1,a1,c0)*E(e2,e3,d1,c1)*F(a0,d0,e2,z0)*G(e3,b1,b2,z1)");
assert(success);
success = exatn::sync(); assert(success);
auto duration = exatn::Timer::timeInSecHR(time_start);
flops = exatn::getTotalFlopCount() - flops;
std::cout << "Time (s) = " << duration << "; GFlop/s = " << flops/duration/1e9 << std::endl << std::flush;
for(int repeat = 0; repeat < 2; ++repeat){
success = exatn::sync(); assert(success);
auto flops = exatn::getTotalFlopCount();
auto time_start = exatn::Timer::timeInSecHR();
success = exatn::evaluateTensorNetwork("MERA1d",
"Z(z0,z1,z2,z3)+=A(a0,a1,a2,z2)*B(b0,b1,b2,z3)*C(c0,c1,a2,b0)*D(d0,d1,a1,c0)*E(e2,e3,d1,c1)*F(a0,d0,e2,z0)*G(e3,b1,b2,z1)");
assert(success);
success = exatn::sync(); assert(success);
auto duration = exatn::Timer::timeInSecHR(time_start);
flops = exatn::getTotalFlopCount() - flops;
std::cout << "Time (s) = " << duration << "; GFlop/s = " << flops/duration/1e9 << std::endl << std::flush;
}
success = exatn::destroyTensor("G"); assert(success);
success = exatn::destroyTensor("F"); assert(success);
......@@ -2031,7 +2032,7 @@ TEST(NumServerTester, neurIPS) {
std::string deriv_tens_name;
for(auto tens = network_ttn_conj->cbegin(); tens != network_ttn_conj->cend(); ++tens){
if(tens->first != 0){
if(tens->first != 0 && tens->second.getRank() == 3){
deriv_tens_name = tens->second.getName();
break;
}
......@@ -2044,20 +2045,23 @@ TEST(NumServerTester, neurIPS) {
}
for(auto net = derivative.begin(); net != derivative.end(); ++net){
success = exatn::initTensorsRnd(*(net->network)); assert(success);
success = exatn::initTensorsRndSync(*(net->network)); assert(success);
}
auto deriv_output_tensor = derivative[0].network->getTensor(0);
success = exatn::createTensorSync("acc",TENS_ELEM_TYPE,deriv_output_tensor->getShape()); assert(success);
success = exatn::initTensor("acc",0.0); assert(success);
success = exatn::initTensorSync("acc",0.0); assert(success);
auto flops = exatn::getTotalFlopCount();
auto time_start = exatn::Timer::timeInSecHR();
success = exatn::evaluate(derivative,exatn::getTensor("acc")); assert(success);
success = exatn::sync(); assert(success);
auto duration = exatn::Timer::timeInSecHR(time_start);
flops = exatn::getTotalFlopCount() - flops;
std::cout << "Time (s) = " << duration << "; GFlop/s = " << flops/duration/1e9 << std::endl << std::flush;
for(int repeat = 0; repeat < 2; ++repeat){
success = exatn::sync(); assert(success);
auto flops = exatn::getTotalFlopCount();
auto time_start = exatn::Timer::timeInSecHR();
success = exatn::evaluate(derivative,exatn::getTensor("acc")); assert(success);
success = exatn::sync(); assert(success);
auto duration = exatn::Timer::timeInSecHR(time_start);
flops = exatn::getTotalFlopCount() - flops;
std::cout << "Time (s) = " << duration << "; GFlop/s = " << flops/duration/1e9 << std::endl << std::flush;
}
success = exatn::destroyTensorSync("acc"); assert(success);
for(auto net = derivative.begin(); net != derivative.end(); ++net){
......@@ -2097,16 +2101,19 @@ TEST(NumServerTester, neurIPS) {
success = exatn::initTensorRnd("I"); assert(success);
success = exatn::initTensorRnd("J"); assert(success);
success = exatn::initTensorRnd("K"); assert(success);
success = exatn::sync(); assert(success);
auto flops = exatn::getTotalFlopCount();
auto time_start = exatn::Timer::timeInSecHR();
success = exatn::evaluateTensorNetwork("ML_MERA","Z(z0,z1,z2)+=A(z0,a1,a2)*B(z1,z2,b2,b3)*C(a1,c1)*D(a2,b2,d2)*E(b3,e1)*F(c1,f1,f2)*G(d2,e1,g2,g3)*H(f1,h1)*I(f2,g2,i2)*J(g3,j1)*K(h1,i2,j1)");
assert(success);
success = exatn::sync(); assert(success);
auto duration = exatn::Timer::timeInSecHR(time_start);
flops = exatn::getTotalFlopCount() - flops;
std::cout << "Time (s) = " << duration << "; GFlop/s = " << flops/duration/1e9 << std::endl << std::flush;
for(int repeat = 0; repeat < 2; ++repeat){
success = exatn::sync(); assert(success);
auto flops = exatn::getTotalFlopCount();
auto time_start = exatn::Timer::timeInSecHR();
success = exatn::evaluateTensorNetwork("ML_MERA",
"Z(z0,z1,z2)+=A(z0,a1,a2)*B(z1,z2,b2,b3)*C(a1,c1)*D(a2,b2,d2)*E(b3,e1)*F(c1,f1,f2)*G(d2,e1,g2,g3)*H(f1,h1)*I(f2,g2,i2)*J(g3,j1)*K(h1,i2,j1)");
assert(success);
success = exatn::sync(); assert(success);
auto duration = exatn::Timer::timeInSecHR(time_start);
flops = exatn::getTotalFlopCount() - flops;
std::cout << "Time (s) = " << duration << "; GFlop/s = " << flops/duration/1e9 << std::endl << std::flush;
}
success = exatn::destroyTensor("K"); assert(success);
success = exatn::destroyTensor("J"); assert(success);
......@@ -2125,6 +2132,7 @@ TEST(NumServerTester, neurIPS) {
//Synchronize:
success = exatn::sync(); assert(success);
exatn::deactivateContrSeqCaching();
exatn::resetLoggingLevel(0,0);
//Grab a beer!
}
......
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment