cd $CAFFE_ROOT/examples/cifar10
./train_quick.sh
-`train_quick.sh` is a simple script, so have a look inside. `GLOG_logtostderr=1` is the google logging flag that prints all the logging messages directly to stderr. The main tool for training is `train_net.bin`, with the solver protobuf text file as its argument.
+`train_quick.sh` is a simple script, so have a look inside. `GLOG_logtostderr=1` is the google logging flag that prints all the logging messages directly to stderr. The main tool for training is `caffe.bin` with the `train` action, and the solver protobuf text file as its argument.
When you run the code, you will see a lot of messages flying by like this:
TOOLS=../../build/tools
-GLOG_logtostderr=1 $TOOLS/train_net.bin \
- cifar10_full_solver.prototxt
+GLOG_logtostderr=1 $TOOLS/caffe.bin train \
+ --solver_proto_file=cifar10_full_solver.prototxt
#reduce learning rate by factor of 10
-GLOG_logtostderr=1 $TOOLS/train_net.bin \
- cifar10_full_solver_lr1.prototxt \
- cifar10_full_iter_60000.solverstate
+GLOG_logtostderr=1 $TOOLS/caffe.bin train \
+ --solver_proto_file=cifar10_full_solver_lr1.prototxt \
+ --resume_point_file=cifar10_full_iter_60000.solverstate
#reduce learning rate by factor of 10
-GLOG_logtostderr=1 $TOOLS/train_net.bin \
- cifar10_full_solver_lr2.prototxt \
- cifar10_full_iter_65000.solverstate
+GLOG_logtostderr=1 $TOOLS/caffe.bin train \
+ --solver_proto_file=cifar10_full_solver_lr2.prototxt \
+ --resume_point_file=cifar10_full_iter_65000.solverstate
TOOLS=../../build/tools
-GLOG_logtostderr=1 $TOOLS/train_net.bin cifar10_quick_solver.prototxt
+GLOG_logtostderr=1 $TOOLS/caffe.bin train \
+ --solver_proto_file=cifar10_quick_solver.prototxt
#reduce learning rate by fctor of 10 after 8 epochs
-GLOG_logtostderr=1 $TOOLS/train_net.bin cifar10_quick_solver_lr1.prototxt cifar10_quick_iter_4000.solverstate
+GLOG_logtostderr=1 $TOOLS/caffe.bin train \
+ --solver_proto_file=cifar10_quick_solver_lr1.prototxt \
+ --resume_point_file=cifar10_quick_iter_4000.solverstate
TOOLS=../../build/tools
-GLOG_logtostderr=1 $TOOLS/train_net.bin \
- imagenet_solver.prototxt caffe_imagenet_train_10000.solverstate
+GLOG_logtostderr=1 $TOOLS/caffe.bin train\
+ --solver_proto_file=imagenet_solver.prototxt \
+ --resume_point_file=caffe_imagenet_train_10000.solverstate
echo "Done."
TOOLS=../../build/tools
-GLOG_logtostderr=1 $TOOLS/train_net.bin alexnet_solver.prototxt
+GLOG_logtostderr=1 $TOOLS/caffe.bin train \
+ --solver_proto_file=alexnet_solver.prototxt
echo "Done."
TOOLS=../../build/tools
-GLOG_logtostderr=1 $TOOLS/train_net.bin imagenet_solver.prototxt
+GLOG_logtostderr=1 $TOOLS/caffe.bin train \
+ --solver_proto_file=imagenet_solver.prototxt
echo "Done."
cd $CAFFE_ROOT/examples/mnist
./train_lenet.sh
-`train_lenet.sh` is a simple script, but here are a few explanations: `GLOG_logtostderr=1` is the google logging flag that prints all the logging messages directly to stderr. The main tool for training is `train_net.bin`, with the solver protobuf text file as its argument.
+`train_lenet.sh` is a simple script, but here are a few explanations: `GLOG_logtostderr=1` is the google logging flag that prints all the logging messages directly to stderr. The main tool for training is `caffe.bin` with action `train`, with the solver protobuf text file as its argument.
When you run the code, you will see a lot of messages flying by like this:
TOOLS=../../build/tools
-GLOG_logtostderr=1 $TOOLS/train_net.bin lenet_solver.prototxt
+GLOG_logtostderr=1 $TOOLS/caffe.bin train \
+ --solver_proto_file=lenet_solver.prototxt
TOOLS=../../build/tools
-GLOG_logtostderr=1 $TOOLS/train_net.bin lenet_consolidated_solver.prototxt
+GLOG_logtostderr=1 $TOOLS/caffe.bin train \
+ --solver_proto_file=lenet_consolidated_solver.prototxt
#!/bin/bash
TOOLS=../../build/tools
-GLOG_logtostderr=1 $TOOLS/train_net.bin mnist_autoencoder_solver.prototxt
+GLOG_logtostderr=1 $TOOLS/caffe.bin train \
+ --solver_proto_file=mnist_autoencoder_solver.prototxt
#include "caffe/common.hpp"
int main(int argc, char** argv) {
- LOG(ERROR) << "Deprecated. Use caffe.bin devicequery "
+ LOG(FATAL) << "Deprecated. Use caffe.bin devicequery "
"[--device_id=0] instead.";
return 0;
}
#include "caffe/caffe.hpp"
int main(int argc, char** argv) {
- LOG(ERROR) << "Deprecated. Use caffe.bin train --solver_proto_file=... "
+ LOG(FATAL) << "Deprecated. Use caffe.bin train --solver_proto_file=... "
"[--pretrained_net_file=...] instead.";
return 0;
}
#include "caffe/caffe.hpp"
int main(int argc, char** argv) {
- LOG(ERROR) << "Deprecated. Use caffe.bin speedtest --net_proto_file=... "
+ LOG(FATAL) << "Deprecated. Use caffe.bin speedtest --net_proto_file=... "
"[--run_iterations=50] [--speedtest_with_gpu] [--device_id=0]";
return 0;
}
#include "caffe/caffe.hpp"
int main(int argc, char** argv) {
- LOG(ERROR) << "Deprecated. Use caffe.bin train --solver_proto_file=... "
+ LOG(FATAL) << "Deprecated. Use caffe.bin train --solver_proto_file=... "
"[--resume_point_file=...] instead.";
return 0;
}