modified: pool_text_to_hdf5_data.py

[gwljf]

change fsg to tsg

[imag2046]

when I run the commad : python s2vt_captioner.py -m s2vt_vgg_rgb , it shows the following errors. I think it's caused by the recurrent , but I don't know how to solve it?

Someone help me? Thanks very much.

---

Setting up LSTM NET [libprotobuf ERROR google/protobuf/text_format.cc:245] Error parsing text-format caffe.NetParameter: 105:19: Message type "caffe.LayerParameter" has no field named "recurrent_param". WARNING: Logging before InitGoogleLogging() is written to STDERR F0628 11:02:25.316423 5425 upgrade_proto.cpp:928] Check failed: ReadProtoFromTextFile(param_file, param) Failed to parse NetParameter file: ./s2vt.words_to_preds.deploy.prototxt *** Check failure stack trace: *** Aborted (core dumped)

[zhengziqiang]

I have the same error, and I found the recurrent_param in the cafffe.proto. So I solved this problem by changing the captioner.py

[Miffy1216]

The model and data cannot be downloaded using 'get_s2vt.sh'. I tried to manually download them but still failed. Can you provide them to me?

[tejus-gupta]

@zhengziqiang What changes did you make to captioner.py to resolve that error?

[Victoria2333]

@gwljf@imag2046@zhengziqiang @Miffy1216 @tejus-gupta Hi,May I ask if you have MSVD Youtube video dataset??? If you do ,would you mind sending it to me??? I'm in desperate need for this but google is forbidden recently, and my email address is [email protected] Thank you for your attention~~~

[vsubhashini]

Fir the MSVD dataset, the train validation and test features (for all frames) are all here: https://www.dropbox.com/sh/whatkfg5mr4dr63/AACKCO3LwSsHK4_GOmHn4oyYa?dl=0

The download_data.sh script downloads the pre-processed video features.

On Oct 21, 2017 10:02 PM, "Victoria2333" [email protected] wrote:

@gwljf https://github.com/gwljf@imag2046@zhengziqiang @Miffy1216 https://github.com/miffy1216 @tejus-gupta https://github.com/tejus-gupta Hi,May I ask if you have MSVD Youtube video dataset??? If you do ,would you mind sending it to me??? I'm in desperate need for this but google is forbidden recently, and my email address is [email protected] Thank you for your attention~~~

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/vsubhashini/caffe/pull/1#issuecomment-338447530, or mute the thread https://github.com/notifications/unsubscribe-auth/AAWOcvarAjfFnI6spxv7gbMyrjgxGWzwks5surBigaJpZM4H_U9M .

[xyIsHere]

@zhengziqiang I have the same error, can you tell me how to fix that problem in captioner.py. Thank you!!!

[dharmisha]

When I run the s2vt_captioner.py I get the following error:

mkldnn_concat_layer.cpp:139] Check failed: dim_src == bottom[i]->shape().size() (4 vs. 3)

I am running on intel caffe version 1.1.0 in cpu only mode.

How can I resolve the same?

[benguftlr]

Hi, when I run the extract_vgg_fetures.py and s2vt_captioner.py I get the following error:

Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)

I couldn't find the exact reason for the error. How can i fix it? Could it be because of the caffe compilation? Thank you for your attention.