It includes a snapshot of the repository in a notebook for exporting to Kaggle for training + submission.
The release here will focus only on supervised training without data transformations. We want to see if the model is trainable using the limited amount of supervised data available to us. It is also good to get a baseline to then see if unsupervised or self-supervised pre-training on the unlabeled data will help.
This is the first release of this project.
It includes a snapshot of the repository in a notebook for exporting to Kaggle for training + submission.
The release here will focus only on supervised training without data transformations. We want to see if the model is trainable using the limited amount of supervised data available to us. It is also good to get a baseline to then see if unsupervised or self-supervised pre-training on the unlabeled data will help.