提交 023baaae authored 作者: Nicolas Ballas's avatar Nicolas Ballas

update warning messages

上级 5ca9ce17
...@@ -100,15 +100,17 @@ def conv2d(input, filters, input_shape=None, image_shape=None, filter_shape=None ...@@ -100,15 +100,17 @@ def conv2d(input, filters, input_shape=None, image_shape=None, filter_shape=None
if len(kargs.keys()) > 0: if len(kargs.keys()) > 0:
warnings.warn(str(kargs.keys()) + warnings.warn(str(kargs.keys()) +
" are not deprecared in " " are now deprecared in "
"`tensor.nnet.abstract_conv.conv2d` interface") "`tensor.nnet.abstract_conv.conv2d` interface"
" and will be ignored.")
if image_shape is not None: if image_shape is not None:
warnings.warn("image_shape is not supported in " warnings.warn("image_shape is not longer supported in "
"`tensor.nnet.abstract_conv.conv2d` interface" "`tensor.nnet.abstract_conv.conv2d` interface"
" use input_shape instead.") " use input_shape instead.")
if input_shape is None: if input_shape is None:
input_shape = image_shape raise ValueError("input_shape and image_shape should not"
" be provided at the same time.")
return abstract_conv2d(input, filters, input_shape, filter_shape, return abstract_conv2d(input, filters, input_shape, filter_shape,
border_mode, subsample=(1, 1)) border_mode, subsample=(1, 1))
......
Markdown 格式
0%
您添加了 0 到此讨论。请谨慎行事。
请先完成此评论的编辑!
注册 或者 后发表评论