multi-task-learning-example

multi-task-learning-example copied to clipboard

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

2

I'm not in the field of deep learning and computer science, but I found this work very interesting. I am confused about what should I do if I want to...

Qianqian-Yang

Log var can become negative and explode

1

The loss function can optimize in a way that keep decreasing the log_var values, which I observe in my experiments. One simple solution is to do torch.abs(log_var). Any thoughts on...

snie2012

Why return torch.mean(loss)?

1

@yaringal Hi, I have a question about your multi-task loss function. Below you return a loss as torch.mean(loss), but if i undersatnd this function correctly, loss is just a single...

edshkim98

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

2

I think this is a lucky demo. When I chenge the data generation code, the optimization will be guided wrongly and the variance prediction is wrong. So I think this...

conan-ux

As described in this paper, the noise(sigma) increases, the respected L(W) decreases. But if we understand sigma as uncertainty of y, maybe it's better for L to be increase with...

caishiqing

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

3

songzeballboy

Sounds like a lucky result comes from a wrong formula deduction

2

I read the paper carefully, the formula in paper is fundamentally wrong. - Under the formula (2) and (3), the probility output has a gaussian distribution. However, the probility can't...

edfall

uncertainty for self-supervised learning

1

@yaringal Thank you for your example, it helps a lot to understand the paper. I am currently use the proposed formula (exp(-log_var)*loss+log_var)) in self-supervised learning with uncertainty estimation. In my...

griffintin

Any way to incorporate these methods in other tasks easily?

7

Hi, thanks for your excellent job! I wonder is there any way to easily incorporate this method into other multi-tasks learning pipeline? I'm still trying to understanding the formulas and...

Jacobew

Hi，thanks for your great works, and i have some questions about the formulation 10 - it says "in the last transition we introduced the explicit simplifying assumption ... which becomes...

neil-yc