yandex-research / rtdl

Research on Tabular Deep Learning: Papers & Packages
Apache License 2.0
888 stars 98 forks source link

怎么运行? #64

Closed Cgetier520990 closed 1 month ago

Cgetier520990 commented 3 months ago

你好,请问这个代码怎么运行?代码好像只提供了模型,没有提供数据。换句话说,这个代码怎么调用表格数据,如果我想给表格的数据分类,我应该怎么做?

Yura52 commented 1 month ago

Sorry for the slow reply, is the help still needed?

Cgetier520990 commented 1 month ago

Thank you for your reply. Well, I still need your help. I can't run your project mainly because there is a lack of training functions. My data set is in CSV format, which contains both text data and numerical data. I want to train with my own data and then do the verification. What should I do?

Yura52 commented 1 month ago

This project only provides a list of papers. Some of the papers have their own Python packages that provide PyTorch modules and end-to-end examples showing how to train them (example). However, there are no high-level training functions in those packages.

Also, there is Skorch -- a package that provides a Scikit-learn wrapper for any PyTorch module, so you can take one of our models and pass it to the wrapper. Though you should not rely on the default hyperparameters in that package, they can be far from optimal.

Does this help?

P.S. The package rtdl is deprecated, so individual packages of the aforementioned papers should be used instead.

Cgetier520990 commented 1 month ago

yeah,I maybe get it, thank you very much.

Yura52 commented 1 month ago

Feel free to reopen the issue if further help is needed