Fix MultiLabelMarginLoss docs (#3836)

zou3519 · soumith · commit 82e39d1231db · 2018-02-06T20:46:32.000-08:00
diff --git a/torch/nn/modules/loss.py b/torch/nn/modules/loss.py
@@ -472,7 +472,7 @@ class MultiLabelMarginLoss(_Loss):
 
     `y` and `x` must have the same size.
 
-    The criterion only considers the first non zero `y[j]` targets.
+    The criterion only considers the first non-negative `y[j]` targets.
 
     This allows for different samples to have variable amounts of target classes
     """
