How to get the weight after training a SeriesNetwork?
5 visualizaciones (últimos 30 días)
Mostrar comentarios más antiguos
Hannah
el 2 de Dic. de 2017
Comentada: Hannah
el 5 de Dic. de 2017
Hello,
I tried to get the weight of my network after training it. So I called getwb(net), but I just got the following error: No appropriate method, property, or field 'numLayers' for class 'SeriesNetwork'. With layers(2).Weights I just got the Weights I started with. Has anybody an idea what I did wrong or what I can do to get the trained weights?
Thanks a lot!
0 comentarios
Respuesta aceptada
Mukul Rao
el 5 de Dic. de 2017
Hello, I work for the MathWorks technical support team. The "getwb" function is only supported by legacy neural network objects. The "SeriesNetwork" object, which is relatively new does not have a documented method for obtaining weights and biases. There is an existing enhancement request in place.
It is possible however, to print the weights and biases associated for each layer.
For example,
net = alexnet
net.Layers(2).Weights %Weights of Convolution layer
net.Layers(2).Bias %Biases of Convolution layer
Más respuestas (0)
Ver también
Categorías
Más información sobre Image Data Workflows en Help Center y File Exchange.
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!