Error in Resume raining neural network 'The output of layer 12 is incompatible with the input expected by layer 13'.

2 visualizzazioni (ultimi 30 giorni)
I am trying to resume training a neural network ( finetuned ResNet-50) using the command trainNetwork but I got this error message :
'The output of layer 12 is incompatible with the input expected by layer 13'
what I have so far is autosaved checkpoint 'convnet_checkpoint__254__2017_12_19__06_54_40.mat'
and I have two questions here :
1- whats the reason behind that error noting that I haven't changed anything in the layers' properties of the network (input/output).
2- basically this is my training command :
my_net = trainNetwork(Traindatasource,lgraph,options);
so, I am a bit confused about how to resume the training using A or B command ( after loading the checkpoint network, passing the data and training options):
A)
trainNetwork(trainingdatastor,net.Layers,options) because I can see the after loading the checkpoint Network its named 'net'
B) trainNetwork(trainingdatastore,my_net.Layers,options)
thanks

Risposta accettata

Joakim Lindblad
Joakim Lindblad il 9 Mar 2018
It's because ResNet is a DAG, and thus the net.Layers in the trainNetwork command is not enough, replace it with
trainNetwork(trainingdatastor,layerGraph(net),options)
  3 Commenti

Accedi per commentare.

Più risposte (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by