# Transition Notes

[toc]

## Scripts
- Training script
- Ready for training both architectures
- Not yet pushed to repo, will wait

## Model

- Models:
- ViT: done
- Swin: running now
- DeiT: in progress
- CLIP: in progress
- custom wrapper made to train just the vision branch of the model
- NOTE this in the write-up.
- There was no appropriate column to act as the text input for this model.
- ResNet18: done
- DenseNet121: done
- ConvNeXt_tiny: done

Training data:
In the pretrained models:
- These labels map to ONE (n=10,796):
- Pet was adopted on the same day as it was listed.
- Pet was adopted between 1 and 7 days (1st week) after being listed.
- Pet was adopted between 8 and 30 days (1st month) after being listed.
- Pet was adopted between 31 and 90 days (2nd & 3rd month) after being listed.
- This label maps to ZERO (n=4,197):
- No adoption after 100 days of being listed. (There are no pets in this dataset that waited between 90 and 100 days).

* Not all pets have images (those are dropped) (n≈341)
* There were rows with nan AdoptionSpeeds () -- only the subsampled bias_dev.csv set. This is a result of the test.csv not having any AdoptionSpeed.