Yongshuo Zong
@yongshuo.bsky.social
PhD student @ University of Edinburgh | https://ys-zong.github.io/
Maybe VQGAN? It's quite useful in quantizing other modalities to discrete tokens to integrate into LLMs for both understanding and generation, but perhaps the main thing is about VQ not GAN...
November 27, 2024 at 7:37 PM
Maybe VQGAN? It's quite useful in quantizing other modalities to discrete tokens to integrate into LLMs for both understanding and generation, but perhaps the main thing is about VQ not GAN...