UniWeTok: An Unified Binary Tokenizer with Codebook Size $\mathit{2^{128}}$ for Unified Multimodal Large Language Model

Benchmark Model Rank Results
image-reconstruction-on-imagenetUniWeTok#6FID: 0.79
image-generation-on-imagenet-256x256UniWeTok-H#26FID: 1.38
image-generation-on-imagenet-256x256UniWeTok-L#60FID: 1.68
image-generation-on-imagenet-256x256UniWeTok-B#127FID: 2.35