Fine-Tuning on Flamingo mini for Image Captioning

#1
by DonaldDaz - opened

Could you provide the code to execute fine tuning of the model on a custom dataset?

Hi there, sorry for the late reply.
I uploaded a training script on a separate branch in the repo: https://github.com/dhansmair/flamingo-mini/tree/training. You could adapt that to your needs.
However it is rather basic and also only works with one GPU.
I would love to set up a script using HuggingFace trainer, but unfortunately I don't have time at the moment.

This comment has been hidden

Sign up or log in to comment