search

FlagOpen / FlagEmbedding

Retrieval and Retrieval-augmented LLMs
MIT License
6.81k stars 491 forks source link

VISTA Train & Evaluation Code #1085

Closed kimwongyuda closed 1 week ago

kimwongyuda commented 2 weeks ago

I'm inspired by your nice work, "VISTA: Visualized Text Embedding For Universal Multi-Modal Retrieval" (Visualized BGE).

I would like to reproduce the model but VISTA project (FlagOpen/FlagEmbedding/visual/) doesn't have training and evaluation code unlike other projects.

It has only modelling code. Could you provide those for reproduction and test?

Thank you.

JUNJIE99 commented 2 weeks ago

Thank you for your interest in our work!

We've made the relevant code available here. If you have any additional questions, please feel free to reach out.