Difference between revisions of "Infrastructure/software/pytorch"

From Nordic Language Processing Laboratory
Jump to: navigation, search
(Usage on Abel)
(Installation on Abel)
Line 65: Line 65:
 
   https://raw.githubusercontent.com/OpenNMT/OpenNMT-py/master/requirements.txt
 
   https://raw.githubusercontent.com/OpenNMT/OpenNMT-py/master/requirements.txt
 
pip install -r /tmp/requirements.txt
 
pip install -r /tmp/requirements.txt
 +
</pre>
 +
 +
 +
= Installation on Taito =
 +
 +
<pre>
 +
module purge
 +
module load cuda/9.0
 +
module load python-env/3.5.3
 +
</pre>
 +
 +
<pre>
 +
cd /proj/nlpl/software
 +
svn co http://svn.nlpl.eu/software/pytorch
 +
virtualenv pytorch/0.4.1
 +
</pre>
 +
 +
Next, we need to create a module definition, in this case
 +
<tt>/proj/nlpl/software/modulefiles/nlpl-pytorch/0.4.1.lua</tt>.
 +
 +
<pre>
 +
module load nlpl-pytorch/0.4.1
 +
pip install --upgrade pip
 +
pip install --upgrade -r /proj/nlpl/software/pytorch/0.4.1/modules.txt
 
</pre>
 
</pre>

Revision as of 19:21, 26 September 2018

Background

In mid-2018 at least, PyTorch is one of the widely used programming frameworks for Deep Learning (with corporate backing by facebook); in and of itself PyTorch is not NLP-specific, but it is a critical dependency for several of the NLPL strands of activity.

Usage on Abel

The module nlpl-pytorch provides a PyTorch installation in a Python 3.5 virtual environment. Besides PyTorch and its dependencies (e.g. NumPy), the virtual environment includes a selection of popular add-on packages, e.g. SciKit-Learn, the Python Data Analysis Library (Pandas), GenSim, and Keras. This installation should support both cpu and gpu nodes on Abel.

module purge
module use -a /projects/nlpl/software/modulefiles
module load nlpl-pytorch

There is a short sample program that test availability of cpu vs. gpu computing devices.

python /projects/nlpl/software/pytorch/0.4.1/test.py

Available Versions

On Abel, as of September 2018, PyTorch 0.4.1 is available.

Installation on Abel

module purge
module load cuda/8.0
module load python3/3.5.0
cd /projects/nlpl/software
mkdir pytorch
virtualenv pytorch/0.4.1

Next, we need to create a module definition, in this case /projects/nlpl/software/modulefiles/nlpl-pytorch/0.4.1.

module load nlpl-pytorch/0.4.1
pip install --upgrade pip
pip install --upgrade numpy pillow six
pip install torch torchvision
wget -O /tmp/requirements.txt \
  https://raw.githubusercontent.com/OpenNMT/OpenNMT-py/master/requirements.txt
pip install -r /tmp/requirements.txt


Installation on Taito

module purge
module load cuda/9.0
module load python-env/3.5.3
cd /proj/nlpl/software
svn co http://svn.nlpl.eu/software/pytorch
virtualenv pytorch/0.4.1

Next, we need to create a module definition, in this case /proj/nlpl/software/modulefiles/nlpl-pytorch/0.4.1.lua.

module load nlpl-pytorch/0.4.1
pip install --upgrade pip
pip install --upgrade -r /proj/nlpl/software/pytorch/0.4.1/modules.txt