Issues

• How can I use the trained model for prediction? is it right to use prediction_model.predict(new_x) ?

  yaringal/multi-task-learning-example#17 · Filter by author Qianqian-Yang opened on Apr 17, 2022
  2

• Why return torch.mean(loss)?

  yaringal/multi-task-learning-example#16 · Filter by author edshkim98 opened on Apr 6, 2021
  1

• Log var can become negative and explode

  yaringal/multi-task-learning-example#14 · Filter by author snie2012 opened on Feb 5, 2021
  1

• Question about the loss

  yaringal/multi-task-learning-example#13 · Filter by author caishiqing opened on Dec 23, 2020

• This is a lucky demo When I change the data generation process, the prediction of the variance is wrong

  yaringal/multi-task-learning-example#12 · Filter by author conan-ux opened on Nov 26, 2020
  2

• uncertainty for self-supervised learning

  yaringal/multi-task-learning-example#11 · Filter by author griffintin opened on Apr 14, 2020
  1

• some questions about formulation 10 in paper

  yaringal/multi-task-learning-example#10 · Filter by author neil-yc opened on Apr 9, 2020

• Sounds like a lucky result comes from a wrong formula deduction

  yaringal/multi-task-learning-example#9 · Filter by author edfall opened on Feb 26, 2020
  2

• Question on relative weights

  yaringal/multi-task-learning-example#8 · Filter by author kmbmjn opened on Nov 28, 2019

• Any way to incorporate these methods in other tasks easily?

  yaringal/multi-task-learning-example#7 · Filter by author Jacobew opened on Sep 22, 2019
  7

• The loss might be nagative value

  yaringal/multi-task-learning-example#6 · Filter by author begeekmyfriend opened on Apr 23, 2019
  11

• MergeLoss with regular item is \log_{sigma} in paper but \log_{sigma}^2 in code

  yaringal/multi-task-learning-example#4 · Filter by author songzeballboy opened on Feb 25, 2019
  3

• about σ

  yaringal/multi-task-learning-example#3 · Filter by author zhangxixi0904 opened on Feb 16, 2019
  1