huji-deep / generative-convacs Goto Github PK

Hello,

After I compiled caffe,

I tried to run "run.py" in/TMM/exp/mnist/ghp_model/train

I got this:

**All required datasets are present.
Generating pre-trained model:

All required datasets are present.
Invalid train plan!

Try python hyper_train.py --help for more information
Error calling hyper_train script
=============== DONE ===============
Invalid train plan!

Try python hyper_train.py --help for more information
Error calling hyper_train script**

I looked up "hyper_train.py" but I am not sure how to fix this error.

Hello,
I already installed Caffe in my computer but I am getting an error.
When I do 'make all' in ~/TMM/deps/siments I got this:

PROTOC src/caffe/proto/caffe.proto
CXX .build_release/src/caffe/proto/caffe.pb.cc
CXX src/caffe/layers/absval_layer.cpp
In file included from ./include/caffe/layer.hpp:12:0,
from ./include/caffe/layers/absval_layer.hpp:7,
from src/caffe/layers/absval_layer.cpp:3:
./include/caffe/util/math_functions.hpp:140:1: error: ‘forceinline’ does not name a type
forceinline device host int8_t caffe_sign(Dtype val) {
^
./include/caffe/util/math_functions.hpp: In function ‘void caffe::caffe_cpu_sign(int, const Dtype*, Dtype*)’:
./include/caffe/util/math_functions.hpp:160:42: error: ‘caffe_sign’ was not declared in this scope
DEFINE_CAFFE_CPU_UNARY_FUNC(sign, y[i] = caffe_sign(x[i]));
^
./include/caffe/util/math_functions.hpp:155:7: note: in definition of macro ‘DEFINE_CAFFE_CPU_UNARY_FUNC’
operation;
^
./include/caffe/util/math_functions.hpp:160:58: error: expected primary-expression before ‘>’ token
DEFINE_CAFFE_CPU_UNARY_FUNC(sign, y[i] = caffe_sign(x[i]));
^
./include/caffe/util/math_functions.hpp:155:7: note: in definition of macro ‘DEFINE_CAFFE_CPU_UNARY_FUNC’
operation;
^
Makefile:572: recipe for target '.build_release/src/caffe/layers/absval_layer.o' failed
make: *** [.build_release/src/caffe/layers/absval_layer.o] Error 1

My OS is ubuntu 16.04.

Hello,

I am trying to do
https://github.com/HUJI-Deep/Generative-ConvACs/tree/master/exp/mnist/dpm

But in the git repository, I do not have "exp/mnist/lenet/train/run.py".
Did "lenet" rename for something else?

Is there any alternative way to redo the experiment?

I am using Ubuntu 16.04 and GTX 1070.

Thank you.