Nick Erickson

@jwmueller my current plan is to add generic support for specifying fold indices / CV splitter (although it will disable repeated bagging when specified). The time-based cross-val will be a...

Hi @jeremyhermann , we are currently working on adding AutoGluon-Forecasting to v0.5 which handles temporal data. As part of this we are working on support for temporal CV splits for...

Thanks for creating an issue! Please provide a full reproducible script along with more information on the machine, such as if it has a GPU.

@ddelange Thanks for looking into this! We want to keep Python 3.7 support for now as it is still widely used. For example, in AWS SageMaker Notebooks, the newest Python...

Great suggestion! I think this is one of the most opaque and confusing aspects of AutoGluon as present. I am in the process of refactoring various aspects of feature generation...

@damirpolat Great! Can you also try (on separate lines): ``` pip install autogluon==0.4.2 pip install protobuf==3.18.1 ``` I believe this should also work, but want to confirm.

@sxjscience and @zhiqiangdon we may want to check if AutoMM Image model can handle very large target values for RMSE, I'm guessing ImagePredictor gets overflow due to the value of...

positive/negative class specification is available in autogluon.tabular but not yet in autogluon.text , this is a good addition to have in a future release. For now, if you set your...

AutoGluon is not currently available via conda, but we plan to add it to conda in future. For now, please follow these instructions to install AutoGluon in a conda environment:...

#263 tracks GPU support for Tabular, I believe the Tabular Neural Network supports GPU at present.