Submission

To submit your trained model/source code/papers to the benchmark, please consider the following:

1. To be eligible to submit your work to the benchmark website, please follow the steps to generate the benchmark dataset, apply preprocessing, create predefine folds at “DATASETS” page.
2. For single-task approaches, we use the predefined five folds and for multitask approaches we randomly split the dataset.
3. Feel free to download our implemented source code and use them as a template for your work.
4. We post your score and model details on the results page.
5. Please send us any available related publications, website, or publicly available code for your model that we can link to your score in results page.
6. Let us know if you want to verify your results. The verified result is trained and tested by our project team members.

Please submit your results by clicking this link.