linting

README.md

@@ -13,7 +13,7 @@
 
 --------------------------------------------------------------------------------
 
-[PyTorch](http://pytorch.org/) (<= 0.4.1) completely lacks autograd support and operations such as sparse sparse matrix multiplication, but is heavily working on improvement (*cf.* [this issue](https://github.com/pytorch/pytorch/issues/9674)).
+[PyTorch](http://pytorch.org/) completely lacks autograd support and operations such as sparse sparse matrix multiplication, but is heavily working on improvement (*cf.* [this issue](https://github.com/pytorch/pytorch/issues/9674)).
 In the meantime, this package consists of a small extension library of optimized sparse matrix operations with autograd support.
 This package currently consists of the following methods:
 

setup.py

@@ -13,10 +13,11 @@ cmdclass = {}
 
 if torch.cuda.is_available():
     ext_modules += [
-        CUDAExtension('spspmm_cuda',
-                      ['cuda/spspmm.cpp', 'cuda/spspmm_kernel.cu'],
-                      extra_link_args=['-lcusparse', '-l', 'cusparse'],
-                      )
+        CUDAExtension(
+            'spspmm_cuda',
+            ['cuda/spspmm.cpp', 'cuda/spspmm_kernel.cu'],
+            extra_link_args=['-lcusparse', '-l', 'cusparse'],
+        )
     ]
     cmdclass['build_ext'] = BuildExtension