New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Potential improvements #10

Open

3 tasks

cosmin-pascaru opened this issue Apr 21, 2019 · 0 comments

Assignees

Member

cosmin-pascaru commented Apr 21, 2019

Heavy regularization of Answer Machine, such that changes from layer to layer will be small, and the final classifier will have an easier job
Scale final classifier losses by depth (the loss at the first layer is quite unimportant, but the loss at the last layer should be very important)

Small things to try:

Calculate the loss for the Confidence Evaluator based only on the first layer (depth 1). With small changes generated by the Q&A Machine, this should translate well to further layers. Needs testing.

cosmin-pascaru assigned simi2525

cosmin-pascaru assigned Tuckle and hhroberthdaniel

simi2525 unassigned Tuckle

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment