Fixes flake8 --select W605 test warnings (#829)

Signed-off-by: asears <asears@users.noreply.github.com>

apex/contrib/optimizers/fused_adam.py

@@ -29,7 +29,7 @@ class FusedAdam(torch.optim.Optimizer):
         use_mt (boolean, optional): use multi tensor apply for lower launch
             latency. (default: False)
 
-    .. _Adam\: A Method for Stochastic Optimization:
+    .. _Adam - A Method for Stochastic Optimization:
         https://arxiv.org/abs/1412.6980
     .. _On the Convergence of Adam and Beyond:
         https://openreview.net/forum?id=ryQu7f-RZ

apex/contrib/optimizers/fused_lamb.py

@@ -54,7 +54,7 @@ class FusedLAMB(torch.optim.Optimizer):
         max_grad_norm (float, optional): value used to clip global grad norm
             (default: 1.0)
 
-    .. _Large Batch Optimization for Deep Learning\: Training BERT in 76 minutes:
+    .. _Large Batch Optimization for Deep Learning - Training BERT in 76 minutes:
         https://arxiv.org/abs/1904.00962
     .. _On the Convergence of Adam and Beyond:
         https://openreview.net/forum?id=ryQu7f-RZ

apex/optimizers/fused_adam.py

@@ -53,7 +53,7 @@ class FusedAdam(torch.optim.Optimizer):
         set_grad_none (bool, optional): whether set grad to None when zero_grad()
             method is called. (default: True)
 
-    .. _Adam\: A Method for Stochastic Optimization:
+    .. _Adam - A Method for Stochastic Optimization:
         https://arxiv.org/abs/1412.6980
     .. _On the Convergence of Adam and Beyond:
         https://openreview.net/forum?id=ryQu7f-RZ

apex/optimizers/fused_lamb.py

@@ -52,7 +52,7 @@ class FusedLAMB(torch.optim.Optimizer):
         max_grad_norm (float, optional): value used to clip global grad norm
             (default: 1.0)
 
-    .. _Large Batch Optimization for Deep Learning\: Training BERT in 76 minutes:
+    .. _Large Batch Optimization for Deep Learning - Training BERT in 76 minutes:
         https://arxiv.org/abs/1904.00962
     .. _On the Convergence of Adam and Beyond:
         https://openreview.net/forum?id=ryQu7f-RZ

apex/optimizers/fused_novograd.py

@@ -58,7 +58,7 @@ class FusedNovoGrad(torch.optim.Optimizer):
         set_grad_none (bool, optional): whether set grad to None when zero_grad()
             method is called. (default: True)
 
-    .. _Jasper\: An End-to-End Convolutional Neural Acoustic Model:
+    .. _Jasper - An End-to-End Convolutional Neural Acoustic Model:
         https://arxiv.org/abs/1904.03288
     .. _On the Convergence of Adam and Beyond:
         https://openreview.net/forum?id=ryQu7f-RZ

apex/reparameterization/__init__.py

@@ -2,7 +2,7 @@ from .weight_norm import WeightNorm
 from .reparameterization import Reparameterization
 
 def apply_weight_norm(module, name='', dim=0, hook_child=True):
-    """
+    r"""
     Applies weight normalization to a parameter in the given module.
     If no parameter is provided, applies weight normalization to all
     parameters in model (except 1-d vectors and scalars).

apex/reparameterization/weight_norm.py

@@ -20,7 +20,7 @@ def _norm(p, dim):
 HALF_TYPES = (torch.cuda.HalfTensor, torch.HalfTensor)
 
 class WeightNorm(Reparameterization):
-    """
+    r"""
     Weight normalization is a reparameterization that decouples the magnitude
     of a weight tensor from its direction. This replaces the parameter specified
     by `name` (e.g. "weight") with two parameters: one specifying the magnitude

requirements_dev.txt

+-r requirements.txt
+flake8>=3.7.9
+Sphinx>=3.0.3
\ No newline at end of file