Replies: 1 comment 3 replies
-
Hi @metliv , Changing to sbert and using the NN-2-Layer is a great idea! Let's try to figure out the root of your issue. As of right now, changing model during the review is not possible, therefore exporting your data and reimporting was the correct choice. Can you run, in the console, the command ^JT |
Beta Was this translation helpful? Give feedback.
3 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi, we are doing the SAFE methode and are in phase 3 now. Is there any way to change the model settings in the project? We wanted to use as a feature extraction technique: Sentence Bert and for the Classfier: fully connected neutral network (2 hidden layers) and we exported a ris-file of the porject with the labeled and und unlabeled records and created a new project with the mentioned settings. We also performed "pip install asreview[all]" to insatll SBert model configuration. I can create the project but when I start reviewig I get the notification "Failed to retrieve new records. Save trained model data before using this function.." How can I fix the problem?
Beta Was this translation helpful? Give feedback.
All reactions