You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by GitBox <gi...@apache.org> on 2017/11/05 15:22:10 UTC
[GitHub] Adesun closed pull request #8548: * [cpp-package] fix cpp example
Adesun closed pull request #8548: * [cpp-package] fix cpp example
URL: https://github.com/apache/incubator-mxnet/pull/8548
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):
diff --git a/cpp-package/example/googlenet.cpp b/cpp-package/example/googlenet.cpp
index ac0585e81a..1b53fd71cb 100644
--- a/cpp-package/example/googlenet.cpp
+++ b/cpp-package/example/googlenet.cpp
@@ -159,8 +159,8 @@ int main(int argc, char const *argv[]) {
train_iter.Reset();
while (train_iter.Next()) {
auto data_batch = train_iter.GetDataBatch();
- args_map["data"] = data_batch.data.Copy(Context::gpu());
- args_map["data_label"] = data_batch.label.Copy(Context::gpu());
+ data_batch.data.CopyTo(&args_map["data"]);
+ data_batch.label.CopyTo(&args_map["data_label"]);
NDArray::WaitAll();
exec->Forward(true);
exec->Backward();
@@ -174,8 +174,8 @@ int main(int argc, char const *argv[]) {
val_iter.Reset();
while (val_iter.Next()) {
auto data_batch = val_iter.GetDataBatch();
- args_map["data"] = data_batch.data.Copy(Context::gpu());
- args_map["data_label"] = data_batch.label.Copy(Context::gpu());
+ data_batch.data.CopyTo(&args_map["data"]);
+ data_batch.label.CopyTo(&args_map["data_label"]);
NDArray::WaitAll();
exec->Forward(false);
NDArray::WaitAll();
diff --git a/cpp-package/example/lenet_with_mxdataiter.cpp b/cpp-package/example/lenet_with_mxdataiter.cpp
index 077f556225..2d1bdb8edb 100644
--- a/cpp-package/example/lenet_with_mxdataiter.cpp
+++ b/cpp-package/example/lenet_with_mxdataiter.cpp
@@ -25,6 +25,7 @@
#include <string>
#include <vector>
#include "mxnet-cpp/MxNetCpp.h"
+#include <chrono>
// Allow IDE to parse the types
#include "../include/mxnet-cpp/op.h"
@@ -89,15 +90,15 @@ int main(int argc, char const *argv[]) {
args_map["fc2_b"] = 0;
auto train_iter = MXDataIter("MNISTIter")
- .SetParam("image", "./train-images-idx3-ubyte")
- .SetParam("label", "./train-labels-idx1-ubyte")
+ .SetParam("image", "./mnist_data/train-images-idx3-ubyte")
+ .SetParam("label", "./mnist_data/train-labels-idx1-ubyte")
.SetParam("batch_size", batch_size)
.SetParam("shuffle", 1)
.SetParam("flat", 0)
.CreateDataIter();
auto val_iter = MXDataIter("MNISTIter")
- .SetParam("image", "./t10k-images-idx3-ubyte")
- .SetParam("label", "./t10k-labels-idx1-ubyte")
+ .SetParam("image", "./mnist_data/t10k-images-idx3-ubyte")
+ .SetParam("label", "./mnist_data/t10k-labels-idx1-ubyte")
.CreateDataIter();
Optimizer* opt = OptimizerRegistry::Find("ccsgd");
@@ -111,35 +112,62 @@ int main(int argc, char const *argv[]) {
auto *exec = lenet.SimpleBind(Context::gpu(), args_map);
auto arg_names = lenet.ListArguments();
+ // Create metrics
+ Accuracy train_acc, val_acc;
+
for (int iter = 0; iter < max_epoch; ++iter) {
- LG << "Epoch: " << iter;
- train_iter.Reset();
- while (train_iter.Next()) {
+ int samples = 0;
+ train_iter.Reset();
+ train_acc.Reset();
+
+ auto tic = chrono::system_clock::now();
+
+ while (train_iter.Next()) {
+ samples += batch_size;
auto data_batch = train_iter.GetDataBatch();
- args_map["data"] = data_batch.data.Copy(Context::gpu());
- args_map["data_label"] = data_batch.label.Copy(Context::gpu());
+
+ data_batch.data.CopyTo(&args_map["data"]);
+ data_batch.label.CopyTo(&args_map["data_label"]);
NDArray::WaitAll();
+
+ // Compute gradients
exec->Forward(true);
exec->Backward();
+
// Update parameters
for (size_t i = 0; i < arg_names.size(); ++i) {
if (arg_names[i] == "data" || arg_names[i] == "data_label") continue;
opt->Update(i, exec->arg_arrays[i], exec->grad_arrays[i]);
}
+
+ // Update metric
+ train_acc.Update(data_batch.label, exec->outputs[0]);
}
+ // one epoch of training is finished
+ auto toc = chrono::system_clock::now();
+ float duration = chrono::duration_cast<chrono::milliseconds>(toc - tic).count() / 1000.0;
+ LG << "Epoch[" << iter << "] " << samples / duration \
+ << " samples/sec " << "Train-Accuracy=" << train_acc.Get();;
+
+ val_iter.Reset();
+ val_acc.Reset();
+
Accuracy acu;
val_iter.Reset();
while (val_iter.Next()) {
auto data_batch = val_iter.GetDataBatch();
- args_map["data"] = data_batch.data.Copy(Context::gpu());
- args_map["data_label"] = data_batch.label.Copy(Context::gpu());
+ data_batch.data.CopyTo(&args_map["data"]);
+ data_batch.label.CopyTo(&args_map["data_label"]);
NDArray::WaitAll();
+
+ // Only forward pass is enough as no gradient is needed when evaluating
exec->Forward(false);
NDArray::WaitAll();
acu.Update(data_batch.label, exec->outputs[0]);
+ val_acc.Update(data_batch.label, exec->outputs[0]);
}
- LG << "Accuracy: " << acu.Get();
+ LG << "Epoch[" << iter << "] Val-Accuracy=" << val_acc.Get();
}
delete exec;
diff --git a/cpp-package/example/mlp_cpu.cpp b/cpp-package/example/mlp_cpu.cpp
index 748c32e8c2..3317f9ed4a 100644
--- a/cpp-package/example/mlp_cpu.cpp
+++ b/cpp-package/example/mlp_cpu.cpp
@@ -106,8 +106,8 @@ int main(int argc, char** argv) {
samples += batch_size;
auto data_batch = train_iter.GetDataBatch();
// Set data and label
- args["X"] = data_batch.data;
- args["label"] = data_batch.label;
+ data_batch.data.CopyTo(&args["X"]);
+ data_batch.label.CopyTo(&args["label"]);
// Compute gradients
exec->Forward(true);
@@ -124,8 +124,8 @@ int main(int argc, char** argv) {
val_iter.Reset();
while (val_iter.Next()) {
auto data_batch = val_iter.GetDataBatch();
- args["X"] = data_batch.data;
- args["label"] = data_batch.label;
+ data_batch.data.CopyTo(&args["X"]);
+ data_batch.label.CopyTo(&args["label"]);
// Forward pass is enough as no gradient is needed when evaluating
exec->Forward(false);
acc.Update(data_batch.label, exec->outputs[0]);
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
With regards,
Apache Git Services