Hi,
Today, when I use the keras I met the problem dropout_score = model.predict_stochastic ,AttributeError: 'Sequential' object has no attribute 'predict_stochastic'. Where or how can I find the description of the parameter about the function?
Thanks!
Peter
predict_stochastic isn't a function anywhere in keras. If you search the code for predict_stochastic there are no results. There is one PR that had a function with that name at one point.
What are you trying to do?
Just to add some context, I think the predict_stochastic function was added by Yarin Gal and Riashat Islam for getting uncertainties thru dropout in neural networks of various kinds. I found the function at https://github.com/Riashat/Active-Learning-Bayesian-Convolutional-Neural-Networks/blob/master/ConvNets/keras/models.py It is also used in Yarin's code at https://github.com/yaringal/BayesianRNN/blob/master/Example/sentiment_lstm_regression.py I ended up at Keras github for LSTM uncertainty estimation because Yarin's page mentions that droput in RNNs has been incorporated in Keras: https://github.com/yaringal/BayesianRNN
This issue has been automatically marked as stale because it has not had recent activity. It will be closed after 30 days if no further activity occurs, but feel free to re-open a closed issue if needed.
Most helpful comment
Just to add some context, I think the predict_stochastic function was added by Yarin Gal and Riashat Islam for getting uncertainties thru dropout in neural networks of various kinds. I found the function at https://github.com/Riashat/Active-Learning-Bayesian-Convolutional-Neural-Networks/blob/master/ConvNets/keras/models.py It is also used in Yarin's code at https://github.com/yaringal/BayesianRNN/blob/master/Example/sentiment_lstm_regression.py I ended up at Keras github for LSTM uncertainty estimation because Yarin's page mentions that droput in RNNs has been incorporated in Keras: https://github.com/yaringal/BayesianRNN