Skip to content

Commit

Permalink
Write long description in PyPI
Browse files Browse the repository at this point in the history
  • Loading branch information
sublee committed Jun 26, 2019
1 parent 76d8a13 commit 9f2658d
Showing 1 changed file with 43 additions and 0 deletions.
43 changes: 43 additions & 0 deletions setup.py
Original file line number Diff line number Diff line change
@@ -1,3 +1,45 @@
"""
torchgpipe
==========
A GPipe_ implementation in PyTorch_.
.. _GPipe: https://arxiv.org/abs/1811.06965
.. _PyTorch: https://pytorch.org/
.. sourcecode:: python
from torchgpipe import GPipe
model = nn.Sequential(a, b, c, d)
model = GPipe(model, balance=[1, 1, 1, 1], chunks=8)
for input in data_loader:
output = model(input)
What is GPipe?
~~~~~~~~~~~~~~
GPipe is a scalable pipeline parallelism library published by Google Brain,
which allows efficient training of large, memory-consuming models. According to
the paper, GPipe can train a 25x larger model by using 8x devices (TPU), and
train a model 3.5x faster by using 4x devices.
`GPipe: Efficient Training of Giant Neural Networks using Pipeline Parallelism
<https://arxiv.org/abs/1811.06965>`_
Google trained AmoebaNet-B with 557M parameters over GPipe. This model has
achieved 84.3% top-1 and 97.0% top-5 accuracy on ImageNet classification
benchmark (the state-of-the-art performance as of May 2019).
Links
~~~~~
- Source Code: https://github.com/kakaobrain/torchgpipe
- Documentation: https://torchgpipe.readthedocs.io/
- Original Paper: https://arxiv.org/abs/1811.06965
"""
from setuptools import setup


Expand All @@ -19,6 +61,7 @@
maintainer='Heungsub Lee, Myungryong Jeong',

description='GPipe for PyTorch',
long_description=__doc__,
keywords='pytorch gpipe',

zip_safe=False,
Expand Down

0 comments on commit 9f2658d

Please sign in to comment.