Even when the Keras config in $HOME/.keras/keras.json is updated to be channels_first data format, this warning is thrown incorrectly while running imdb_cnn example -
'MXNet Backend performs best withchannels_firstformat. Using ' 'channels_lastwill significantly reduce performance due to the ' 'Transpose operations. ' 'For performance improvement, please use this API' 'keras.utils.to_channels_first(x_input)' 'to transformchannels_lastdata tochannels_firstformat and ' 'also please change theimage_data_formatinkeras.jsonto ' 'channels_first.' 'Note:x_inputis a Numpy tensor or a list of Numpy tensor' 'Refer to: ' 'https://github.com/awslabs/keras-apache-mxnet/tree/master/docs/mxnet_backend/performance_guide.md',
This needs to be updated to use K.data_format instead of taking it from the parameter
Even when the Keras config in
$HOME/.keras/keras.json
is updated to bechannels_first
data format, this warning is thrown incorrectly while runningimdb_cnn
example -'MXNet Backend performs best with
channels_firstformat. Using ' '
channels_lastwill significantly reduce performance due to the ' 'Transpose operations. ' 'For performance improvement, please use this API' '
keras.utils.to_channels_first(x_input)' 'to transform
channels_lastdata to
channels_firstformat and ' 'also please change the
image_data_formatin
keras.jsonto ' '
channels_first.' 'Note:
x_inputis a Numpy tensor or a list of Numpy tensor' 'Refer to: ' 'https://github.com/awslabs/keras-apache-mxnet/tree/master/docs/mxnet_backend/performance_guide.md',
This needs to be updated to use K.data_format instead of taking it from the parameter