Ravin Kohli comments

Results 38 comments of


                                            Ravin Kohli

[bug] There might be some memory leak

probably we should follow the recommendations and findings in #180

Refactor autoPyTorch/pipeline/components/preprocessing/tabular_preprocessing/XXX/init.py

It seems like this is a replica of #267. So I'll close that issue and lets carry it from here.

Initial feature encoder 'OneHotEncoder' is not allowed to use in time-series forecasting task

> Thanks for the reporting! We will add another function to change the illegal values to the default configuration values and add it to the next release ASAP. I think...

time-series config json file missing when running example code

Hi, we have fixed this issue in the latest release. You can install it using `pip install autoPyTorch==0.2.1`. I am closing this issue for now.

Subsampling vs feature selection?

For reference, autogluon has implemented feature pruning [here](https://github.com/awslabs/autogluon/pull/1274)

[feat] Add an adaptive subsampler

@nabenabe0928 as we now have #444, do you think we can close this pull request?

Dropout application

Hey, as far as I can tell, dropout_shape was not a hyperparameter but a parameter that depended on the network shape in the master branch. Similarly, it has exactly the...

Dropout application

This will also be fixed in #358

ValueError: Expected parameter df (Tensor of shape (32, 168, 1)) of distribution Chi2() to satisfy the constraint GreaterThan(lower_bound=0.0), but found invalid values

Hi, we have fixed this issue in the latest release. You can install it using `pip install autoPyTorch==0.2.1`. I am closing this issue for now.

Greedy Portfolio on version 0.1.0

This portfolio would be released with the next major release. Till then, we will continue using the portflolio from version 0.0.2