DropoutUncertaintyCaffeModels
DropoutUncertaintyCaffeModels copied to clipboard
Dropout As A Bayesian Approximation: Code
Hi! First of all, thanks for open-sourcing the experiments. I'm currently trying to reproduce the results using a different framework. I noticed that in the definition of the convolutional networks...
Shouldn't the `sample_weight_test` parameter be uncommented in dropout networks (eg. in https://github.com/yaringal/DropoutUncertaintyCaffeModels/blob/master/cifar10_uncertainty/lenet_all_dropout_deploy.prototxt)?
Just a small issue: `message DropoutParameter` in `src/caffe/proto/caffe.proto` is not in line 471 anymore (line 691 in current version of Caffe).
This would be helpful as I'm trying to use for a research project. Specifically, I'm hoping for a CNN trained with a large benchmark image set (ILSVRC or COCO...etc.). Any...