Update.
authorFrancois Fleuret <francois@fleuret.org>
Sat, 15 Apr 2017 23:16:48 +0000 (01:16 +0200)
committerFrancois Fleuret <francois@fleuret.org>
Sat, 15 Apr 2017 23:16:48 +0000 (01:16 +0200)
README.md

index b58970b..82dcf74 100644 (file)
--- a/README.md
+++ b/README.md
@@ -17,12 +17,12 @@ torch.ByteTensor generate_sequence(bool pulling,
 where
 
  * `pulling` indicates if one of the shape should be pulled upward,
- * `nb_sequences` is the number of sequences to generate
- * `nb_images` is the number of frames per sequence
- * `image_height` and `image_width` is the individual image size
- * `nb_shapes` is the number of rectangular shapes in the simulation
- * `random_shape_size` states if the shapes should be of different sizes
- * `random_colors` states if they should be of different colors
+ * `nb_sequences` is the number of sequences to generate,
+ * `nb_images` is the number of frames per sequence,
+ * `image_height` and `image_width` is the individual image size,
+ * `nb_shapes` is the number of rectangular shapes in the simulation,
+ * `random_shape_size` states if the shapes should be of different sizes,
+ * `random_colors` states if they should be of different colors.
 
 The returned ByteTensor has five dimensions: