search

YuanJianhao508 / RAG-Driver

A Multi-Modal Large Language Model with Retrieval-augmented In-context Learning capacity designed for generalisable and explainable end-to-end driving
Apache License 2.0
63 stars 4 forks source link

Model checkpoints #10

Closed halikhani closed 3 days ago

halikhani commented 2 months ago

Hi, very interesting work! I was wondering when we can access model checkpoints. Thank you !

YuanJianhao508 commented 1 month ago

Hi @halikhani , thank you for your interest in our work! We are currently cleaning and gathering everything to release and will update you once it is ready!

halikhani commented 1 month ago

Thanks a lot! actually I was able to fine-tune the model, but there is a big gap between the results I got and the results provided in the paper. What kind of scaling did you use since as far as I know, the range for metrics like METEOR is 0-1 and for CIDEr is 0-10. I assume all the numbers in your evaluation section for these metrics are multiplied by 100 from the original metrics. Please let me know if my assumption is correct. Thank you!

YuanJianhao508 commented 3 days ago

Hi @halikhani , thank you for asking! Yes, for the metric we scale it by 100 just for the purpose for demonstrating the result. Thank you very much!