Skip to content

avlaskin/dev-world-2017-coreml

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 

Repository files navigation

These are support links for the talk:

Follow Andrew Ng on YouTube:

Good learning resources for Machine Learning and Deep Learning:

How to set up Keras on Mac (assuming that you have Python installed):

curl https://bootstrap.pypa.io/get-pip.py > get-pip.py
python get-pip.py
sudo pip install tensorflow# for tensorflow-gpu see below
sudo pip install keras# coremltools now supports keras 2.x
sudo pip install coremltools
cd ~/.keras# Here we need to check if keras backend is tensorflow
cat keras.json

And your ~/.keras/keras.json should look something like this:

{
    "image_dim_ordering": "tf",
    "epsilon": 1e-07,
    "floatx": "float32",
    "backend": "tensorflow"
}

To use (external or internal for old Macs) Nvidia GPU we need to install following:

  1. The NVIDIA drivers associated with CUDA.
  2. CUDA Toolkit 8.0 [v8.0.61]
  3. cuDNN v5.1 for CUDA 8.0.[v5.1] (this needs Nvidia developer account) - (Should be installed with DYLD_LIBRARY_PATH set)
  4. GPU card with CUDA compute capability 3.0 or higher.
  5. Set up DYLD_LIBRARY_PATH to be exported with cuda/lib and cuDNN/lib paths included.
  6. After that we can install tensorflow-gpu

sudo pip install tensorflow-gpu

Enjoy!

This repo has an example of Keras Neural network that does learn XOR function. Although training set is excessive (takes all possible cases) - it is a good starting point for understanding on how you can do Keras model for iOS.

Contacts:

About

Research notes for Dev World 2017 on talk "Smart Machines are waking up in your pockets"

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages