Boostrap Scripts to Install and Test Setup for Deep Learning Applications

README.md

Boostrap Scripts to Install and Test Setup for Deep Learning Applications

This gist is simply a set of scripts to install and test everything you need to get started with Deep Learning. This includes:

• TensorFlow
• Theano
• Keras
• CUDNN
• CUDA Toolkit
• Misc. linux Tools like tmux.

How to use

• Assuming you already have a newly created EC2 GPU instance, and are able to SSH into it.

• Copy deep_learning_bootstrap.sh, test_install.sh and theano_test.py on the machine.

• Run chmod +x *.sh to make the .sh files executable

• Run sudo ./deep_learning_bootstrap.sh and go have a cup of tea while it installs everything for you

  • If a pink screen pops up mentioning "A new version of /boot/grub/menu.lst is available", choose "Keep local version" and select OK.

• When the message to restart system comes up, just restart with sudo shutdown -r 0

• After restart, run ./test_install.sh

  • If everything has gone correct, then you will see output matching the expected output (in green). You are all set to get started!
  • In case of errors, please ping me at @navinpai on twitter or email at navin[at]fiftheye[dot]in

Version

1.0b

Thanks and Have Fun!

Navin "M@dMAx" Pai (@navinpai)

deep_learning_bootstrap.sh

#!/bin/bash

apt-get update
apt-get -y dist-upgrade
apt-get install -y gcc g++ gfortran build-essential git wget tmux linux-image-extra-`uname -r` linux-generic linux-image-generic libopenblas-dev python-dev python-pip python-nose python-numpy python-scipy

# Bleeding edge Theano
pip install --upgrade --no-deps git+git://github.com/Theano/Theano.git

# Install Tensorflow
sudo pip install six --upgrade --target="/usr/lib/python2.7/dist-packages"
pip install --upgrade https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow-0.10.0rc0-cp27-none-linux_x86_64.whl

# Keras v.1.0.7 
pip install --upgrade --no-deps keras==1.0.7

# Scikit-Learn
pip install -U scikit-learn

# CUDA
curl -fsSL http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1404/x86_64/cuda-repo-ubuntu1404_7.0-28_amd64.deb -O
dpkg -i cuda-repo-ubuntu1404_7.0-28_amd64.deb
apt-get update
apt-get install -y cuda
echo -e "\nexport PATH=/usr/local/cuda/bin:$PATH\n\nexport LD_LIBRARY_PATH=/usr/local/cuda/lib64\n\nexport CUDA_HOME=/usr/local/cuda\n\nexport PS1='\[\033[0;32m\]\u@\[\033[0;33m\]D33P_L34RN:\[\033[36m\]\W\[\033[0m\] \$ '" >> .bashrc

# CUDNN 
curl -fsSL http://developer.download.nvidia.com/compute/redist/cudnn/v4/cudnn-7.0-linux-x64-v4.0-prod.tgz -O
tar xvzf cudnn-7.0-linux-x64-v4.0-prod.tgz 
cp cuda/include/cudnn.h /usr/local/cuda/include
cp cuda/lib64/libcudnn* /usr/local/cuda/lib64
chmod a+r /usr/local/cuda/include/cudnn.h /usr/local/cuda/lib64/libcudnn*

# Make Theano use GPU by default
echo -e "\n[global]\nfloatX=float32\ndevice=gpu\n[mode]=FAST_RUN\n\n[nvcc]\nfastmath=True\n\n[cuda]\nroot=/usr/local/cuda" >> ~/.theanorc

# Reboot system and run test_install.sh
echo -e '\E[33;92m'"Reboot System (sudo shutdown -r 0) and run ./test_install.sh"; tput sgr0

test_install.sh

#!/bin/bash

# Test CUDA install
echo -e '\E[33;92m'"TEST CUDA INSTALL"; tput sgr0
cuda-install-samples-7.5.sh ~
cd NVIDIA_CUDA-7.5_Samples/1_Utilities/deviceQuery
make
./deviceQuery
echo -e '\E[33;92m'"SHOULD PRINT Result = PASS ABOVE THIS LINE"; tput sgr0

# Test Theano GPU usage
echo -e '\E[33;92m'""; tput sgr0
cd ~
python theano_test.py
echo -e '\E[33;92m'"SHOULD PRINT Used the gpu ABOVE THIS LINE"; tput sgr0

# Test Theano CDNN usage
echo -e '\E[33;92m'"TEST Theano CDNN USAGE"; tput sgr0
python -c 'from theano.sandbox.cuda.dnn import version; print(version())'
echo -e '\E[33;92m'"SHOULD PRINT (4007, 4007) ABOVE THIS LINE"; tput sgr0

# Run MNIST on TensorFlow
echo -e '\E[33;92m'"TEST TENSORFLOW (Might take a while...)"; tput sgr0
python -m tensorflow.models.image.mnist.convolutional
echo -e '\E[33;92m'"Should Print Validation and Test Error numbers ABOVE THIS LINE"; tput sgr0

echo -e '\E[1;33;92m'"If No errors Seen, then you are done! Have fun! :) - M@dMAx"; tput sgr0

theano_test.py

from theano import function, config, shared, sandbox
import theano.tensor as T
import numpy
import time

vlen = 10 * 30 * 768  # 10 x #cores x # threads per core
iters = 1000

rng = numpy.random.RandomState(22)
x = shared(numpy.asarray(rng.rand(vlen), config.floatX))
f = function([], T.exp(x))
print(f.maker.fgraph.toposort())
t0 = time.time()
for i in range(iters):
    r = f()
t1 = time.time()
print("Looping %d times took %f seconds" % (iters, t1 - t0))
print("Result is %s" % (r,))
if numpy.any([isinstance(x.op, T.Elemwise) for x in f.maker.fgraph.toposort()]):
    print('Used the cpu')
else:
    print('Used the gpu')

If you get the following error on running python -m tensorflow.models.image.mnist.convolutional:

Traceback (most recent call last): File "/usr/lib/python2.7/runpy.py", line 151, in _run_module_as_main mod_name, loader, code, fname = _get_module_details(mod_name) File "/usr/lib/python2.7/runpy.py", line 101, in _get_module_details loader = get_loader(mod_name) File "/usr/lib/python2.7/pkgutil.py", line 464, in get_loader return find_loader(fullname) File "/usr/lib/python2.7/pkgutil.py", line 474, in find_loader for importer in iter_importers(fullname): File "/usr/lib/python2.7/pkgutil.py", line 430, in iter_importers __import__(pkg) File "/usr/local/lib/python2.7/dist-packages/tensorflow/__init__.py", line 23, in <module> from tensorflow.python import * File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/__init__.py", line 98, in <module> from tensorflow.python.platform import test File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/platform/test.py", line 63, in <module> from tensorflow.python.framework import test_util File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/test_util.py", line 43, in <module> from tensorflow.python.platform import googletest File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/platform/googletest.py", line 32, in <module> from tensorflow.python.platform import benchmark # pylint: disable=unused-import File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/platform/benchmark.py", line 112, in <module> class Benchmark(six.with_metaclass(_BenchmarkRegistrar, object)): File "/usr/lib/python2.7/dist-packages/six.py", line 617, in with_metaclass return meta("NewBase", bases, {}) File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/platform/benchmark.py", line 107, in __new__ if not newclass.is_abstract(): AttributeError: type object 'NewBase' has no attribute 'is_abstract'

run

sudo pip install six --upgrade --target="/usr/lib/python2.7/dist-packages"
and then try again