Tong Xiao

We followed the Faster-RCNN's approach to 1. resizing image so that the shortest side is 600px 2. if the longest side is over 1000px, we resize the image again so...

Thanks very much for your advice. I will try my best to figure it out how to make the pretrained models compatible with tensorflow. Need some days though.

@sde123 Could you please attached the full error report? Does this happen when compiling or running the code? It's also better to open up a new issue for it.

@sde123 The mpi version is too old. To install a newer version, please download the source from [here](https://www.open-mpi.org/software/ompi/v1.10/downloads/openmpi-1.10.7.tar.gz), then ```bash tar xf openmpi-1.10.7.tar.gz cd openmpi-1.10.7 ./configure --with-cuda=/usr/local/cuda --enable-mpi-thread-multiple make -j8...

@sde123 This means the training failed. Have you compiled the caffe successfully with the new openmpi?

@sde123 Yeah, I know the Makefile.config. I mean are there any errors when you run `make` to compile the caffe? A tip: when copy and paste something, please use the...

@sde123 Ok, I assume you have compiled caffe successfully. Could you please comment out [this line](https://github.com/Cysu/dgd_person_reid/blob/master/scripts/exp_individually.sh#L20) and run it again? It will reveal the true error.

Could you please check the outputs of the following commands ```bash ldd external/caffe/build/tools/caffe | grep mpi which mpi mpirun --version echo $LD_LIBRARY_PATH ```

@sde123 Sorry for my typo, it should be `which mpirun` not `which mpi`. Also what's the output of ```bash ldd external/caffe/build/tools/caffe | grep mpi ``` Please use ``` (the tilde...

@sde123 The command should be ```bash ldd external/caffe/build/tools/caffe | grep mpi ``` Note that it is `grep mpi` not `grep mpirun`. By the way, what's your GPU memory size?