What did I do today

  • Stratified labeling (single person either in train or test dataset)
  • Served transformation package

What I learned from others

  • Use of augumentation like TTA, cutmix
  • Early stopping
  • Model is almost always about data -> importance of augmentation

What I should do tomorrow

  • data rearrange -> place wrongly labeled data at the right place manually
  • f1 loss
  • try lots of models

Peer Session

Reflections

- Make a plan always!