PaliGemma (PG) is a family of Vision Language Models from Google. It uses SigLIP as the vision encoder, and the Gemma family of models as it language counterpart.
Note
I and Ritwik Raha have covered SigLIP in depth in our blog Choosing Between SigLIP and CLIP for Language Image Pretraining if you wanted to read about it.
PaliGemma is great for fine tuning purposes. In this repository we collect some examples of fine tuning the PG family of models.
Models hosted on Hugging Face Hub:
In the script provided we have used the tuxemon
dataset, from the diffusers team. The dataset comprises of images of tuxemons (a spin off of pokemons)
and their captions.
| Before Fine Tuning | After Fine Tuning |
|---|---|
 |
 |
While I could not find a document that provides pointers to train the model on a detection dataset, diving in the official big vision space made it really clear. Taking inspiration from the space, I have create a script to format any object detection dataset (here the dataset is based on the coco format) to the format PaliGemma is trained on.
You can find the dataset creation script here: create_od_dataset.py.
After the dataset is created run the fine tuning script object_detection_ft.py
and run the model.
| Before Fine Tuning | After Fine Tuning |
|---|---|
 |
 |
Find more information in the count intersection readme.
If you like our work and would use it please cite us
@misc{github_repository,
author = {Aritra Roy Gosthipaty, Ritwik Raha},
title = {ft-pali-gemma},
publisher = {{GitHub}(https://github.com)},
howpublished = {\url{https://github.com/ariG23498/ft-pali-gemma/edit/main/README.md}},
year = {2024}
}