1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
|
#ifndef _TASMANIAN_BENCHMARK_LOADNEEDED_HPP
#define _TASMANIAN_BENCHMARK_LOADNEEDED_HPP
#include "benchCommon.hpp"
bool benchmark_loadneeded(std::deque<std::string> &args){
if (args.size() < 11) return false;
// report the test parameters to reference later
cout << "loadneeded";
for(auto &s : args) cout << " " << s;
auto grid_family = getGridFamily(args);
if (grid_family == GridFamily::none) return false;
int num_dimensions, num_outputs, num_depth, order, iteratons, num_jumps, device;
TypeDepth dtype;
TypeOneDRule rule;
TypeAcceleration acc;
auto riter = readEntries(args.begin(), num_dimensions, num_outputs, num_depth, dtype, rule, order, iteratons, num_jumps, acc, device);
std::string flavor = checkFlavor(riter, args.end());
auto extra = extractWeightsLimits(grid_family, num_dimensions, dtype, riter, args.end());
auto make_grid = getLambdaMakeGrid(grid_family, num_dimensions, num_outputs, num_depth, dtype, rule, order, extra);
for(int jump = 0; jump < num_jumps; jump++){
auto grid = make_grid();
//grid.printStats(); // uncomment to make sure the right grid is constructed
if (jump == 0) cout << " (points: " << grid.getNumPoints() << ")" << endl;
grid.enableAcceleration(acc, device);
if (flavor != "auto")
grid.favorSparseAcceleration((flavor == "sparse"));
auto values = getGenericModel((size_t) grid.getNumDimensions(),
(size_t) grid.getNumOutputs(),
grid.getNeededPoints());
cout << setw(7) << testMethod<DryRun>(iteratons, [&](int)->void{ grid.loadNeededPoints(values); }) << endl;
num_outputs *= 2;
}
return true;
}
#endif
|
