Neural-style: No Output

[tegarkurniawan]

I did try and run but no output image
code :
th neural_style.lua -gpu -1 -style_image examples/inputs/picasso_selfport1907.jpg -content_image examples/inputs/brad_pitt.jpg -output_image profile.png -model_file models/VGG_ILSVRC_19_layers.caffemodel -proto_file models/VGG_ILSVRC_19_layers_deploy.prototxt -content_weight 10 -style_weight 250 -image_size 250 -print_iter 1

output code:
[libprotobuf WARNING google/protobuf/io/coded_stream.cc:505] Reading dangerously large protocol message. If the message turns out to be larger than 1073741824 bytes, parsing will be halted for security reasons. To increase the limit (or to disable these warnings), see CodedInputStream::SetTotalBytesLimit() in google/protobuf/io/coded_stream.h.
[libprotobuf WARNING google/protobuf/io/coded_stream.cc:78] The total number of bytes read was 574671192
Successfully loaded models/VGG_ILSVRC_19_layers.caffemodel
conv1_1: 64 3 3 3
conv1_2: 64 64 3 3
conv2_1: 128 64 3 3
conv2_2: 128 128 3 3
conv3_1: 256 128 3 3
conv3_2: 256 256 3 3
conv3_3: 256 256 3 3
conv3_4: 256 256 3 3
conv4_1: 512 256 3 3
conv4_2: 512 512 3 3
conv4_3: 512 512 3 3
conv4_4: 512 512 3 3
conv5_1: 512 512 3 3
conv5_2: 512 512 3 3
conv5_3: 512 512 3 3
conv5_4: 512 512 3 3
fc6: 1 1 25088 4096
fc7: 1 1 4096 4096
fc8: 1 1 4096 1000
Setting up style layer 2 : relu1_1
Setting up style layer 7 : relu2_1
Setting up style layer 12 : relu3_1
Setting up style layer 21 : relu4_1
Setting up content layer 23 : relu4_2
Setting up style layer 30 : relu5_1
Running optimization with L-BFGS
Iteration 1 / 1000
Content 1 loss: 6830778.125000
Style 1 loss: 156138.580322
Style 2 loss: 20665453.125000
Style 3 loss: 6607033.203125
Style 4 loss: 252967109.375000
Style 5 loss: 21495.241165
Total loss: 287248007.649612
creating recyclable direction/step/history buffers
Iteration 2 / 1000
Content 1 loss: 6830768.125000
Style 1 loss: 156138.580322
Style 2 loss: 20665453.125000
Style 3 loss: 6607033.203125
Style 4 loss: 252967109.375000
Style 5 loss: 21495.241165
Total loss: 287247997.649612
Iteration 3 / 1000
Content 1 loss: 6330779.375000
Style 1 loss: 139395.187378
Style 2 loss: 15094379.882812
Style 3 loss: 5382623.046875
Style 4 loss: 204632218.750000
Style 5 loss: 20202.121735
Total loss: 231599598.363800
Iteration 4 / 1000
Content 1 loss: 380513320.000000
Style 1 loss: 6452613.769531
Style 2 loss: 4886064000.000000
Style 3 loss: 8913999000.000000
Style 4 loss: 1254110464000.000000
Style 5 loss: 594051125.000000
Total loss: 1268891544058.769531
Iteration 5 / 1000
Content 1 loss: 6332839.375000
Style 1 loss: 139340.881348
Style 2 loss: 15071003.906250
Style 3 loss: 5368807.128906
Style 4 loss: 203264937.500000
Style 5 loss: 20150.247574
Total loss: 230197079.039078

whether it can be without being able to install CUDA ?

[3DTOPO]

Did you let it get to integration 1000? By default it saves an intermediate image every 100 iterations - so that is the soonest you would see any images output unless you use the save_iter flag.

[tegarkurniawan]

I'm only up to 5 because of the integration of a very long , how long I have to wait until 1000 integration ?

[3DTOPO]

Unless you use your GPU, it will take a very long time. You use the save_iter flag to specify that you want images written maybe every 5 or 10 iterations (for example: -save_iter 10) - but it typically takes hundreds of iterations to look like anything useful.

[tegarkurniawan]

I have added -save_inter 100 but old
code :
th neural_style.lua -gpu -1 -style_image examples/inputs/picasso_selfport1907.jpg -content_image examples/inputs/brad_pitt.jpg -output_image profile.png -model_file models/VGG_ILSVRC_19_layers.caffemodel -proto_file models/VGG_ILSVRC_19_layers_deploy.prototxt -content_weight 10 -style_weight 250 -image_size 250 -save_iter 100

how long to get the picture output ?

[3DTOPO]

100 is the default - so you didn't change anything and you will have to wait until iteration 100 before you see any images output.

[tegarkurniawan]

100 is max interation ?
if I change my code:
th neural_style.lua -gpu -1 -style_image examples/inputs/picasso_selfport1907.jpg -content_image examples/inputs/brad_pitt.jpg -output_image profile.png -model_file models/VGG_ILSVRC_19_layers.caffemodel -proto_file models/VGG_ILSVRC_19_layers_deploy.prototxt -content_weight 10 -style_weight 250 -image_size 250 -save_iter 10

do i have to wait till 10 iteration for get output picture ?

[3DTOPO]

100 is max interation ?

By default 1000 is max. By default it writes one image every 100 iterations.

do i have to wait till 10 iteration for get output picture ?

Correct.

[tegarkurniawan]

thanks i already get the output image , but I felt too long to get a perfect result , do you have any suggestions?

[3DTOPO]

It often takes some experimenting to tune results. Also, check the wiki and there are some helpful threads here.

You'll want to get a supported GPU - it is like ~100 times faster.