search

hclhkbu / dlbench

Benchmarking State-of-the-Art Deep Learning Software Tools
http://dlbench.comp.hkbu.edu.hk/
MIT License
170 stars 47 forks source link

Provide instructions for running benchmarks #6

Open yaroslavvb opened 7 years ago

yaroslavvb commented 7 years ago

It would be useful to have a readme to reproduce the benchmarks. For instance I found that for running cifar multi-gpu resnet on TF I needed to do pip install scikit-image

Also, it seems to look for CIFAR image data inside of /home/<username>/data/tensorflow/cifar10 for some reason, so I also had to do

cp -R <cifar datadir created when running official CIFAR example >/cifar-10-batches-bin /home/<username>/data/tensorflow/cifar10

FreemanX commented 7 years ago

We are working docs of this repo. As for the data, a symbolic link will work just fine if you want to save some space.