add setup.py and uploaded the library to PyPI so that it can be installed throught pip

Author: yaozhewei

README.md

@@ -25,9 +25,17 @@ Please contact us if you have a related paper and we would be glad to add it to
 
 
 ## Usage
-Please first clone the PyHessian library to your local system:
+### Install from Pip
+You can install the library from pip
+```
+pip install pyhessian
+```
+
+### Install from source
+You can also compile the library from source
 ```
 git clone https://github.com/amirgholami/PyHessian.git
+python setup.py install
 ```
 
 Before running the Hessian code, we need a (pre-trained) NN model. Here, we provide a training file to train ResNet20 model on Cifar-10 dataset:

setup.py

@@ -0,0 +1,17 @@
+from setuptools import setup
+
+setup(
+    # Needed to silence warnings (and to be a worthwhile package)
+    name='PyHessian',
+    url='https://github.com/amirgholami/PyHessian',
+    author='Zhewei Yao, Amir Gholami',
+    author_email='[email protected], [email protected]',
+    # Needed to actually package something
+    packages=['pyhessian'],
+    # Needed for dependencies
+    install_requires=['numpy', 'torch'],
+    # Others
+    version='0.0.1',
+    license='MIT',
+    description='Pytorch library for second-order based analysis and training of Neural Networks',
+)