combining "domain adaptation" technique and chemprop #917
SoodabehGhaffari
started this conversation in
General
Replies: 1 comment
-
To my knowledge, there aren't any chemprop papers about domain adaptation, but an internet search of "domain adaptation chemistry" comes up with a couple of general results. Whether it will improve model performance to initialize the weights from another model trained on a different task is usually hard to predict. Best to try it and see what the empirical results are. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hello,
Previously, I asked a question about transfer learning and comparing it with modeling from scratch. I tried the pretrained models in chemprop to build a model for a related task. However, I realized the transfer learning model is not alway better than the models from scratch. Recently, I came across with a new method called "domain adaptation". I was wondering if chemprop has the feature to do "domain adaptation" and if there is any related paper in this regard.
I really appreciate your help.
Best Regards
Soo
Beta Was this translation helpful? Give feedback.
All reactions