search

IDEA-Research / GroundingDINO

[ECCV 2024] Official implementation of the paper "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection"
https://arxiv.org/abs/2303.05499
Apache License 2.0
6.9k stars 699 forks source link

LoRA finetuning #172

Open ThiagoViek opened 1 year ago

ThiagoViek commented 1 year ago

In the LLMs field, people are going after foundation models and finetuning them to their own tasks using Low-rank Adaptation. Is it possible to leverage LoRA for finetuning Grounding DINO as well?

rentainhe commented 1 year ago

In the LLMs field, people are going after foundation models and finetuning them to their own tasks using Low-rank Adaptation. Is it possible to leverage LoRA for finetuning Grounding DINO as well?

Sure, I think it's a good way to adapt Grounding-DINO in other long-tail domain task

Nina0109 commented 1 year ago

@ThiagoViek Hi, I'm also interested in finetuning grounding dino to fit some personal tasks. Have you ever tried lora on grounding dino? Looking forward for further discussion.

THUeeY commented 7 months ago

@ThiagoViek Remarkable ideas! Have you tried some experiments? Expecting for some feedback.

ahan-2000 commented 5 months ago

https://github.com/BeileiCui/SurgicalDINO

This is something I found, but I am not sure if this is useful.